Daily Papers

Recent advancements in image and video synthesis have opened up new promise in generative games. One particularly intriguing application is transforming characters from anime films into interactive, playable entities. This allows players to immerse themselves in the dynamic anime world as their favorite characters for life simulation through language instructions. Such games are defined as infinite game since they eliminate predetermined boundaries and fixed gameplay rules, where players can interact with the game world through open-ended language and experience ever-evolving storylines and environments. Recently, a pioneering approach for infinite anime life simulation employs large language models (LLMs) to translate multi-turn text dialogues into language instructions for image generation. However, it neglects historical visual context, leading to inconsistent gameplay. Furthermore, it only generates static images, failing to incorporate the dynamics necessary for an engaging gaming experience. In this work, we propose AnimeGamer, which is built upon Multimodal Large Language Models (MLLMs) to generate each game state, including dynamic animation shots that depict character movements and updates to character states, as illustrated in Figure 1. We introduce novel action-aware multimodal representations to represent animation shots, which can be decoded into high-quality video clips using a video diffusion model. By taking historical animation shot representations as context and predicting subsequent representations, AnimeGamer can generate games with contextual consistency and satisfactory dynamics. Extensive evaluations using both automated metrics and human evaluations demonstrate that AnimeGamer outperforms existing methods in various aspects of the gaming experience. Codes and checkpoints are available at https://github.com/TencentARC/AnimeGamer.

We have witnessed impressive advances in video action understanding. Increased dataset sizes, variability, and computation availability have enabled leaps in performance and task diversification. Current systems can provide coarse- and fine-grained descriptions of video scenes, extract segments corresponding to queries, synthesize unobserved parts of videos, and predict context. This survey comprehensively reviews advances in uni- and multi-modal action understanding across a range of tasks. We focus on prevalent challenges, overview widely adopted datasets, and survey seminal works with an emphasis on recent advances. We broadly distinguish between three temporal scopes: (1) recognition tasks of actions observed in full, (2) prediction tasks for ongoing partially observed actions, and (3) forecasting tasks for subsequent unobserved action. This division allows us to identify specific action modeling and video representation challenges. Finally, we outline future directions to address current shortcomings.

Despite an exciting new wave of multimodal machine learning models, current approaches still struggle to interpret the complex contextual relationships between the different modalities present in videos. Going beyond existing methods that emphasize simple activities or objects, we propose a new model-agnostic approach for generating detailed textual descriptions that captures multimodal video information. Our method leverages the extensive knowledge learnt by large language models, such as GPT-3.5 or Llama2, to reason about textual descriptions of the visual and aural modalities, obtained from BLIP-2, Whisper and ImageBind. Without needing additional finetuning of video-text models or datasets, we demonstrate that available LLMs have the ability to use these multimodal textual descriptions as proxies for ``sight'' or ``hearing'' and perform zero-shot multimodal classification of videos in-context. Our evaluations on popular action recognition benchmarks, such as UCF-101 or Kinetics, show these context-rich descriptions can be successfully used in video understanding tasks. This method points towards a promising new research direction in multimodal classification, demonstrating how an interplay between textual, visual and auditory machine learning models can enable more holistic video understanding.

Human action understanding is crucial for the advancement of multimodal systems. While recent developments, driven by powerful large language models (LLMs), aim to be general enough to cover a wide range of categories, they often overlook the need for more specific capabilities. In this work, we address the more challenging task of Fine-grained Action Recognition (FAR), which focuses on detailed semantic labels within shorter temporal duration (e.g., "salto backward tucked with 1 turn"). Given the high costs of annotating fine-grained labels and the substantial data needed for fine-tuning LLMs, we propose to adopt semi-supervised learning (SSL). Our framework, SeFAR, incorporates several innovative designs to tackle these challenges. Specifically, to capture sufficient visual details, we construct Dual-level temporal elements as more effective representations, based on which we design a new strong augmentation strategy for the Teacher-Student learning paradigm through involving moderate temporal perturbation. Furthermore, to handle the high uncertainty within the teacher model's predictions for FAR, we propose the Adaptive Regulation to stabilize the learning process. Experiments show that SeFAR achieves state-of-the-art performance on two FAR datasets, FineGym and FineDiving, across various data scopes. It also outperforms other semi-supervised methods on two classical coarse-grained datasets, UCF101 and HMDB51. Further analysis and ablation studies validate the effectiveness of our designs. Additionally, we show that the features extracted by our SeFAR could largely promote the ability of multimodal foundation models to understand fine-grained and domain-specific semantics.

We present Palm, a solution to the Long-Term Action Anticipation (LTA) task utilizing vision-language and large language models. Given an input video with annotated action periods, the LTA task aims to predict possible future actions. We hypothesize that an optimal solution should capture the interdependency between past and future actions, and be able to infer future actions based on the structure and dependency encoded in the past actions. Large language models have demonstrated remarkable commonsense-based reasoning ability. Inspired by that, Palm chains an image captioning model and a large language model. It predicts future actions based on frame descriptions and action labels extracted from the input videos. Our method outperforms other participants in the EGO4D LTA challenge and achieves the best performance in terms of action prediction. Our code is available at https://github.com/DanDoge/Palm

Visual-Language Models (VLMs) have significantly advanced action video recognition. Supervised by the semantics of action labels, recent works adapt the visual branch of VLMs to learn video representations. Despite the effectiveness proved by these works, we believe that the potential of VLMs has yet to be fully harnessed. In light of this, we exploit the semantic units (SU) hiding behind the action labels and leverage their correlations with fine-grained items in frames for more accurate action recognition. SUs are entities extracted from the language descriptions of the entire action set, including body parts, objects, scenes, and motions. To further enhance the alignments between visual contents and the SUs, we introduce a multi-region module (MRA) to the visual branch of the VLM. The MRA allows the perception of region-aware visual features beyond the original global feature. Our method adaptively attends to and selects relevant SUs with visual features of frames. With a cross-modal decoder, the selected SUs serve to decode spatiotemporal video representations. In summary, the SUs as the medium can boost discriminative ability and transferability. Specifically, in fully-supervised learning, our method achieved 87.8\% top-1 accuracy on Kinetics-400. In K=2 few-shot experiments, our method surpassed the previous state-of-the-art by +7.1% and +15.0% on HMDB-51 and UCF-101, respectively.

This paper introduces a video dataset of spatio-temporally localized Atomic Visual Actions (AVA). The AVA dataset densely annotates 80 atomic visual actions in 430 15-minute video clips, where actions are localized in space and time, resulting in 1.58M action labels with multiple labels per person occurring frequently. The key characteristics of our dataset are: (1) the definition of atomic visual actions, rather than composite actions; (2) precise spatio-temporal annotations with possibly multiple annotations for each person; (3) exhaustive annotation of these atomic actions over 15-minute video clips; (4) people temporally linked across consecutive segments; and (5) using movies to gather a varied set of action representations. This departs from existing datasets for spatio-temporal action recognition, which typically provide sparse annotations for composite actions in short video clips. We will release the dataset publicly. AVA, with its realistic scene and action complexity, exposes the intrinsic difficulty of action recognition. To benchmark this, we present a novel approach for action localization that builds upon the current state-of-the-art methods, and demonstrates better performance on JHMDB and UCF101-24 categories. While setting a new state of the art on existing datasets, the overall results on AVA are low at 15.6% mAP, underscoring the need for developing new approaches for video understanding.

Vision-Language Navigation (VLN) is a challenging task that requires an embodied agent to perform action-level modality alignment, i.e., make instruction-asked actions sequentially in complex visual environments. Most existing VLN agents learn the instruction-path data directly and cannot sufficiently explore action-level alignment knowledge inside the multi-modal inputs. In this paper, we propose modAlity-aligneD Action PrompTs (ADAPT), which provides the VLN agent with action prompts to enable the explicit learning of action-level modality alignment to pursue successful navigation. Specifically, an action prompt is defined as a modality-aligned pair of an image sub-prompt and a text sub-prompt, where the former is a single-view observation and the latter is a phrase like ''walk past the chair''. When starting navigation, the instruction-related action prompt set is retrieved from a pre-built action prompt base and passed through a prompt encoder to obtain the prompt feature. Then the prompt feature is concatenated with the original instruction feature and fed to a multi-layer transformer for action prediction. To collect high-quality action prompts into the prompt base, we use the Contrastive Language-Image Pretraining (CLIP) model which has powerful cross-modality alignment ability. A modality alignment loss and a sequential consistency loss are further introduced to enhance the alignment of the action prompt and enforce the agent to focus on the related prompt sequentially. Experimental results on both R2R and RxR show the superiority of ADAPT over state-of-the-art methods.

A key challenge with procedure planning in instructional videos lies in how to handle a large decision space consisting of a multitude of action types that belong to various tasks. To understand real-world video content, an AI agent must proficiently discern these action types (e.g., pour milk, pour water, open lid, close lid, etc.) based on brief visual observation. Moreover, it must adeptly capture the intricate semantic relation of the action types and task goals, along with the variable action sequences. Recently, notable progress has been made via the integration of diffusion models and visual representation learning to address the challenge. However, existing models employ rudimentary mechanisms to utilize task information to manage the decision space. To overcome this limitation, we introduce a simple yet effective enhancement - a masked diffusion model. The introduced mask acts akin to a task-oriented attention filter, enabling the diffusion/denoising process to concentrate on a subset of action types. Furthermore, to bolster the accuracy of task classification, we harness more potent visual representation learning techniques. In particular, we learn a joint visual-text embedding, where a text embedding is generated by prompting a pre-trained vision-language model to focus on human actions. We evaluate the method on three public datasets and achieve state-of-the-art performance on multiple metrics. Code is available at https://github.com/ffzzy840304/Masked-PDPP.

Understanding human activity is a crucial yet intricate task in egocentric vision, a field that focuses on capturing visual perspectives from the camera wearer's viewpoint. While traditional methods heavily rely on representation learning trained on extensive video data, there exists a significant limitation: obtaining effective video representations proves challenging due to the inherent complexity and variability in human activities.Furthermore, exclusive dependence on video-based learning may constrain a model's capability to generalize across long-tail classes and out-of-distribution scenarios. In this study, we introduce a novel approach for long-term action anticipation using language models (LALM), adept at addressing the complex challenges of long-term activity understanding without the need for extensive training. Our method incorporates an action recognition model to track previous action sequences and a vision-language model to articulate relevant environmental details. By leveraging the context provided by these past events, we devise a prompting strategy for action anticipation using large language models (LLMs). Moreover, we implement Maximal Marginal Relevance for example selection to facilitate in-context learning of the LLMs. Our experimental results demonstrate that LALM surpasses the state-of-the-art methods in the task of long-term action anticipation on the Ego4D benchmark. We further validate LALM on two additional benchmarks, affirming its capacity for generalization across intricate activities with different sets of taxonomies. These are achieved without specific fine-tuning.

We consider the task of temporal human action localization in lifestyle vlogs. We introduce a novel dataset consisting of manual annotations of temporal localization for 13,000 narrated actions in 1,200 video clips. We present an extensive analysis of this data, which allows us to better understand how the language and visual modalities interact throughout the videos. We propose a simple yet effective method to localize the narrated actions based on their expected duration. Through several experiments and analyses, we show that our method brings complementary information with respect to previous methods, and leads to improvements over previous work for the task of temporal action localization.

Multimodal self-supervised learning is getting more and more attention as it allows not only to train large networks without human supervision but also to search and retrieve data across various modalities. In this context, this paper proposes a self-supervised training framework that learns a common multimodal embedding space that, in addition to sharing representations across different modalities, enforces a grouping of semantically similar instances. To this end, we extend the concept of instance-level contrastive learning with a multimodal clustering step in the training pipeline to capture semantic similarities across modalities. The resulting embedding space enables retrieval of samples across all modalities, even from unseen datasets and different domains. To evaluate our approach, we train our model on the HowTo100M dataset and evaluate its zero-shot retrieval capabilities in two challenging domains, namely text-to-video retrieval, and temporal action localization, showing state-of-the-art results on four different datasets.

Interactive and embodied tasks pose at least two fundamental challenges to existing Vision & Language (VL) models, including 1) grounding language in trajectories of actions and observations, and 2) referential disambiguation. To tackle these challenges, we propose an Embodied MultiModal Agent (EMMA): a unified encoder-decoder model that reasons over images and trajectories, and casts action prediction as multimodal text generation. By unifying all tasks as text generation, EMMA learns a language of actions which facilitates transfer across tasks. Different to previous modular approaches with independently trained components, we use a single multitask model where each task contributes to goal completion. EMMA performs on par with similar models on several VL benchmarks and sets a new state-of-the-art performance (36.81% success rate) on the Dialog-guided Task Completion (DTC), a benchmark to evaluate dialog-guided agents in the Alexa Arena

Spatio-temporal grounding describes the task of localizing events in space and time, e.g., in video data, based on verbal descriptions only. Models for this task are usually trained with human-annotated sentences and bounding box supervision. This work addresses this task from a multimodal supervision perspective, proposing a framework for spatio-temporal action grounding trained on loose video and subtitle supervision only, without human annotation. To this end, we combine local representation learning, which focuses on leveraging fine-grained spatial information, with a global representation encoding that captures higher-level representations and incorporates both in a joint approach. To evaluate this challenging task in a real-life setting, a new benchmark dataset is proposed providing dense spatio-temporal grounding annotations in long, untrimmed, multi-action instructional videos for over 5K events. We evaluate the proposed approach and other methods on the proposed and standard downstream tasks showing that our method improves over current baselines in various settings, including spatial, temporal, and untrimmed multi-action spatio-temporal grounding.

Due to the resource-intensive nature of training vision-language models on expansive video data, a majority of studies have centered on adapting pre-trained image-language models to the video domain. Dominant pipelines propose to tackle the visual discrepancies with additional temporal learners while overlooking the substantial discrepancy for web-scaled descriptive narratives and concise action category names, leading to less distinct semantic space and potential performance limitations. In this work, we prioritize the refinement of text knowledge to facilitate generalizable video recognition. To address the limitations of the less distinct semantic space of category names, we prompt a large language model (LLM) to augment action class names into Spatio-Temporal Descriptors thus bridging the textual discrepancy and serving as a knowledge base for general recognition. Moreover, to assign the best descriptors with different video instances, we propose Optimal Descriptor Solver, forming the video recognition problem as solving the optimal matching flow across frame-level representations and descriptors. Comprehensive evaluations in zero-shot, few-shot, and fully supervised video recognition highlight the effectiveness of our approach. Our best model achieves a state-of-the-art zero-shot accuracy of 75.1% on Kinetics-600.

Action localization networks are often structured as a feature encoder sub-network and a localization sub-network, where the feature encoder learns to transform an input video to features that are useful for the localization sub-network to generate reliable action proposals. While some of the encoded features may be more useful for generating action proposals, prior action localization approaches do not include any attention mechanism that enables the localization sub-network to attend more to the more important features. In this paper, we propose a novel attention mechanism, the Class Semantics-based Attention (CSA), that learns from the temporal distribution of semantics of action classes present in an input video to find the importance scores of the encoded features, which are used to provide attention to the more useful encoded features. We demonstrate on two popular action detection datasets that incorporating our novel attention mechanism provides considerable performance gains on competitive action detection models (e.g., around 6.2% improvement over BMN action detection baseline to obtain 47.5% mAP on the THUMOS-14 dataset), and a new state-of-the-art of 36.25% mAP on the ActivityNet v1.3 dataset. Further, the CSA localization model family which includes BMN-CSA, was part of the second-placed submission at the 2021 ActivityNet action localization challenge. Our attention mechanism outperforms prior self-attention modules such as the squeeze-and-excitation in action detection task. We also observe that our attention mechanism is complementary to such self-attention modules in that performance improvements are seen when both are used together.

Prompt-based learning has been demonstrated as a compelling paradigm contributing to large language models' tremendous success (LLMs). Inspired by their success in language tasks, existing research has leveraged LLMs in embodied instruction following and task planning. However, not much attention has been paid to embodied tasks with multimodal prompts, combining vision signals with text descriptions. This type of task poses a major challenge to robots' capability to understand the interconnection and complementarity between vision and language signals. In this work, we introduce an effective framework that learns a policy to perform robot manipulation with multimodal prompts from multi-task expert trajectories. Our methods consist of a two-stage training pipeline that performs inverse dynamics pretraining and multi-task finetuning. To facilitate multimodal understanding, we design our multimodal prompt encoder by augmenting a pretrained LM with a residual connection to the visual input and model the dependencies among action dimensions. Empirically, we evaluate the efficacy of our method on the VIMA-BENCH and establish a new state-of-the-art (10% improvement in success rate). Moreover, we demonstrate that our model exhibits remarkable in-context learning ability.

Human beings possess the capability to multiply a melange of multisensory cues while actively exploring and interacting with the 3D world. Current multi-modal large language models, however, passively absorb sensory data as inputs, lacking the capacity to actively interact with the objects in the 3D environment and dynamically collect their multisensory information. To usher in the study of this area, we propose MultiPLY, a multisensory embodied large language model that could incorporate multisensory interactive data, including visual, audio, tactile, and thermal information into large language models, thereby establishing the correlation among words, actions, and percepts. To this end, we first collect Multisensory Universe, a large-scale multisensory interaction dataset comprising 500k data by deploying an LLM-powered embodied agent to engage with the 3D environment. To perform instruction tuning with pre-trained LLM on such generated data, we first encode the 3D scene as abstracted object-centric representations and then introduce action tokens denoting that the embodied agent takes certain actions within the environment, as well as state tokens that represent the multisensory state observations of the agent at each time step. In the inference time, MultiPLY could generate action tokens, instructing the agent to take the action in the environment and obtain the next multisensory state observation. The observation is then appended back to the LLM via state tokens to generate subsequent text or action tokens. We demonstrate that MultiPLY outperforms baselines by a large margin through a diverse set of embodied tasks involving object retrieval, tool use, multisensory captioning, and task decomposition.

Actions are about how we interact with the environment, including other people, objects, and ourselves. In this paper, we propose a novel multi-modal Holistic Interaction Transformer Network (HIT) that leverages the largely ignored, but critical hand and pose information essential to most human actions. The proposed "HIT" network is a comprehensive bi-modal framework that comprises an RGB stream and a pose stream. Each of them separately models person, object, and hand interactions. Within each sub-network, an Intra-Modality Aggregation module (IMA) is introduced that selectively merges individual interaction units. The resulting features from each modality are then glued using an Attentive Fusion Mechanism (AFM). Finally, we extract cues from the temporal context to better classify the occurring actions using cached memory. Our method significantly outperforms previous approaches on the J-HMDB, UCF101-24, and MultiSports datasets. We also achieve competitive results on AVA. The code will be available at https://github.com/joslefaure/HIT.

What makes good video representations for video understanding, such as anticipating future activities, or answering video-conditioned questions? While earlier approaches focus on end-to-end learning directly from video pixels, we propose to revisit text-based representations, such as discrete action labels, or free-form video captions, which are interpretable and can be directly consumed by large language models (LLMs). Intuitively, different video understanding tasks may require representations that are complementary and at different granularities. To this end, we propose versatile action models (Vamos), a learning framework powered by a large language model as the "reasoner", and can flexibly leverage visual embeddings, action labels, and free-form descriptions extracted from videos as its input. We evaluate Vamos on four complementary video understanding benchmarks, Ego4D, Next-QA, IntentQA, and EgoSchema, on its capability to model temporal dynamics, encode visual history, and perform reasoning. Surprisingly, we observe that text-based representations consistently achieve competitive performance on all benchmarks, and that visual embeddings provide marginal or no performance improvement, demonstrating the effectiveness of text-based video representation in the LLM era. We perform extensive ablation study and qualitative analysis to support our observations, and achieve state-of-the-art performance on three benchmarks.

We introduce Image-GOal Representations (IGOR), aiming to learn a unified, semantically consistent action space across human and various robots. Through this unified latent action space, IGOR enables knowledge transfer among large-scale robot and human activity data. We achieve this by compressing visual changes between an initial image and its goal state into latent actions. IGOR allows us to generate latent action labels for internet-scale video data. This unified latent action space enables the training of foundation policy and world models across a wide variety of tasks performed by both robots and humans. We demonstrate that: (1) IGOR learns a semantically consistent action space for both human and robots, characterizing various possible motions of objects representing the physical interaction knowledge; (2) IGOR can "migrate" the movements of the object in the one video to other videos, even across human and robots, by jointly using the latent action model and world model; (3) IGOR can learn to align latent actions with natural language through the foundation policy model, and integrate latent actions with a low-level policy model to achieve effective robot control. We believe IGOR opens new possibilities for human-to-robot knowledge transfer and control.

Temporal action localization (TAL), which involves recognizing and locating action instances, is a challenging task in video understanding. Most existing approaches directly predict action classes and regress offsets to boundaries, while overlooking the discrepant importance of each frame. In this paper, we propose an Action Sensitivity Learning framework (ASL) to tackle this task, which aims to assess the value of each frame and then leverage the generated action sensitivity to recalibrate the training procedure. We first introduce a lightweight Action Sensitivity Evaluator to learn the action sensitivity at the class level and instance level, respectively. The outputs of the two branches are combined to reweight the gradient of the two sub-tasks. Moreover, based on the action sensitivity of each frame, we design an Action Sensitive Contrastive Loss to enhance features, where the action-aware frames are sampled as positive pairs to push away the action-irrelevant frames. The extensive studies on various action localization benchmarks (i.e., MultiThumos, Charades, Ego4D-Moment Queries v1.0, Epic-Kitchens 100, Thumos14 and ActivityNet1.3) show that ASL surpasses the state-of-the-art in terms of average-mAP under multiple types of scenarios, e.g., single-labeled, densely-labeled and egocentric.

Delving into the realm of egocentric vision, the advancement of referring video object segmentation (RVOS) stands as pivotal in understanding human activities. However, existing RVOS task primarily relies on static attributes such as object names to segment target objects, posing challenges in distinguishing target objects from background objects and in identifying objects undergoing state changes. To address these problems, this work proposes a novel action-aware RVOS setting called ActionVOS, aiming at segmenting only active objects in egocentric videos using human actions as a key language prompt. This is because human actions precisely describe the behavior of humans, thereby helping to identify the objects truly involved in the interaction and to understand possible state changes. We also build a method tailored to work under this specific setting. Specifically, we develop an action-aware labeling module with an efficient action-guided focal loss. Such designs enable ActionVOS model to prioritize active objects with existing readily-available annotations. Experimental results on VISOR dataset reveal that ActionVOS significantly reduces the mis-segmentation of inactive objects, confirming that actions help the ActionVOS model understand objects' involvement. Further evaluations on VOST and VSCOS datasets show that the novel ActionVOS setting enhances segmentation performance when encountering challenging circumstances involving object state changes. We will make our implementation available at https://github.com/ut-vision/ActionVOS.

Due to the distinctive characteristics of sensors, each modality exhibits unique physical properties. For this reason, in the context of multi-modal action recognition, it is important to consider not only the overall action content but also the complementary nature of different modalities. In this paper, we propose a novel network, named Modality Mixer (M-Mixer) network, which effectively leverages and incorporates the complementary information across modalities with the temporal context of actions for action recognition. A key component of our proposed M-Mixer is the Multi-modal Contextualization Unit (MCU), a simple yet effective recurrent unit. Our MCU is responsible for temporally encoding a sequence of one modality (e.g., RGB) with action content features of other modalities (e.g., depth and infrared modalities). This process encourages M-Mixer network to exploit global action content and also to supplement complementary information of other modalities. Furthermore, to extract appropriate complementary information regarding to the given modality settings, we introduce a new module, named Complementary Feature Extraction Module (CFEM). CFEM incorporates sepearte learnable query embeddings for each modality, which guide CFEM to extract complementary information and global action content from the other modalities. As a result, our proposed method outperforms state-of-the-art methods on NTU RGB+D 60, NTU RGB+D 120, and NW-UCLA datasets. Moreover, through comprehensive ablation studies, we further validate the effectiveness of our proposed method.

Image-based visual-language (I-VL) pre-training has shown great success for learning joint visual-textual representations from large-scale web data, revealing remarkable ability for zero-shot generalisation. This paper presents a simple but strong baseline to efficiently adapt the pre-trained I-VL model, and exploit its powerful ability for resource-hungry video understanding tasks, with minimal training. Specifically, we propose to optimise a few random vectors, termed as continuous prompt vectors, that convert video-related tasks into the same format as the pre-training objectives. In addition, to bridge the gap between static images and videos, temporal information is encoded with lightweight Transformers stacking on top of frame-wise visual features. Experimentally, we conduct extensive ablation studies to analyse the critical components. On 10 public benchmarks of action recognition, action localisation, and text-video retrieval, across closed-set, few-shot, and zero-shot scenarios, we achieve competitive or state-of-the-art performance to existing methods, despite optimising significantly fewer parameters.

Existing action recognition methods are typically actor-specific due to the intrinsic topological and apparent differences among the actors. This requires actor-specific pose estimation (e.g., humans vs. animals), leading to cumbersome model design complexity and high maintenance costs. Moreover, they often focus on learning the visual modality alone and single-label classification whilst neglecting other available information sources (e.g., class name text) and the concurrent occurrence of multiple actions. To overcome these limitations, we propose a new approach called 'actor-agnostic multi-modal multi-label action recognition,' which offers a unified solution for various types of actors, including humans and animals. We further formulate a novel Multi-modal Semantic Query Network (MSQNet) model in a transformer-based object detection framework (e.g., DETR), characterized by leveraging visual and textual modalities to represent the action classes better. The elimination of actor-specific model designs is a key advantage, as it removes the need for actor pose estimation altogether. Extensive experiments on five publicly available benchmarks show that our MSQNet consistently outperforms the prior arts of actor-specific alternatives on human and animal single- and multi-label action recognition tasks by up to 50%. Code is made available at https://github.com/mondalanindya/MSQNet.

We present ALFRED (Action Learning From Realistic Environments and Directives), a benchmark for learning a mapping from natural language instructions and egocentric vision to sequences of actions for household tasks. ALFRED includes long, compositional tasks with non-reversible state changes to shrink the gap between research benchmarks and real-world applications. ALFRED consists of expert demonstrations in interactive visual environments for 25k natural language directives. These directives contain both high-level goals like "Rinse off a mug and place it in the coffee maker." and low-level language instructions like "Walk to the coffee maker on the right." ALFRED tasks are more complex in terms of sequence length, action space, and language than existing vision-and-language task datasets. We show that a baseline model based on recent embodied vision-and-language tasks performs poorly on ALFRED, suggesting that there is significant room for developing innovative grounded visual language understanding models with this benchmark.

While open-source multi-modal language models perform well on simple question answering tasks, they often fail on complex questions that require multiple capabilities, such as fine-grained recognition, visual grounding, and reasoning, and that demand multi-step solutions. We present TACO, a family of multi-modal large action models designed to improve performance on such complex, multi-step, and multi-modal tasks. During inference, TACO produces chains-of-thought-and-action (CoTA), executes intermediate steps by invoking external tools such as OCR, depth estimation and calculator, then integrates both the thoughts and action outputs to produce coherent responses. To train TACO, we create a large dataset of over 1M synthetic CoTA traces generated with GPT-4o and Python programs. We then experiment with various data filtering and mixing techniques and obtain a final subset of 293K high-quality CoTA examples. This dataset enables TACO to learn complex reasoning and action paths, surpassing existing models trained on instruction tuning data with only direct answers. Our model TACO outperforms the instruction-tuned baseline across 8 benchmarks, achieving a 3.6% improvement on average, with gains of up to 15% in MMVet tasks involving OCR, mathematical reasoning, and spatial reasoning. Training on high-quality CoTA traces sets a new standard for complex multi-modal reasoning, highlighting the need for structured, multi-step instruction tuning in advancing open-source mutli-modal models' capabilities.

This paper focuses on temporal localization of actions in untrimmed videos. Existing methods typically train classifiers for a pre-defined list of actions and apply them in a sliding window fashion. However, activities in the wild consist of a wide combination of actors, actions and objects; it is difficult to design a proper activity list that meets users' needs. We propose to localize activities by natural language queries. Temporal Activity Localization via Language (TALL) is challenging as it requires: (1) suitable design of text and video representations to allow cross-modal matching of actions and language queries; (2) ability to locate actions accurately given features from sliding windows of limited granularity. We propose a novel Cross-modal Temporal Regression Localizer (CTRL) to jointly model text query and video clips, output alignment scores and action boundary regression results for candidate clips. For evaluation, we adopt TaCoS dataset, and build a new dataset for this task on top of Charades by adding sentence temporal annotations, called Charades-STA. We also build complex sentence queries in Charades-STA for test. Experimental results show that CTRL outperforms previous methods significantly on both datasets.

We propose a self-supervised algorithm to learn representations from egocentric video data. Recently, significant efforts have been made to capture humans interacting with their own environments as they go about their daily activities. In result, several large egocentric datasets of interaction-rich multi-modal data have emerged. However, learning representations from videos can be challenging. First, given the uncurated nature of long-form continuous videos, learning effective representations require focusing on moments in time when interactions take place. Second, visual representations of daily activities should be sensitive to changes in the state of the environment. However, current successful multi-modal learning frameworks encourage representation invariance over time. To address these challenges, we leverage audio signals to identify moments of likely interactions which are conducive to better learning. We also propose a novel self-supervised objective that learns from audible state changes caused by interactions. We validate these contributions extensively on two large-scale egocentric datasets, EPIC-Kitchens-100 and the recently released Ego4D, and show improvements on several downstream tasks, including action recognition, long-term action anticipation, and object state change classification.

We aim to automatically identify human action reasons in online videos. We focus on the widespread genre of lifestyle vlogs, in which people perform actions while verbally describing them. We introduce and make publicly available the WhyAct dataset, consisting of 1,077 visual actions manually annotated with their reasons. We describe a multimodal model that leverages visual and textual information to automatically infer the reasons corresponding to an action presented in the video.

Visual tempo, which describes how fast an action goes, has shown its potential in supervised action recognition. In this work, we demonstrate that visual tempo can also serve as a self-supervision signal for video representation learning. We propose to maximize the mutual information between representations of slow and fast videos via hierarchical contrastive learning (VTHCL). Specifically, by sampling the same instance at slow and fast frame rates respectively, we can obtain slow and fast video frames which share the same semantics but contain different visual tempos. Video representations learned from VTHCL achieve the competitive performances under the self-supervision evaluation protocol for action recognition on UCF-101 (82.1\%) and HMDB-51 (49.2\%). Moreover, comprehensive experiments suggest that the learned representations are generalized well to other downstream tasks including action detection on AVA and action anticipation on Epic-Kitchen. Finally, we propose Instance Correspondence Map (ICM) to visualize the shared semantics captured by contrastive learning.

With the release of large-scale motion datasets with textual annotations, the task of establishing a robust latent space for language and 3D human motion has recently witnessed a surge of interest. Methods have been proposed to convert human motion and texts into features to achieve accurate correspondence between them. Despite these efforts to align language and motion representations, we claim that the temporal element is often overlooked, especially for compound actions, resulting in chronological inaccuracies. To shed light on the temporal alignment in motion-language latent spaces, we propose Chronologically Accurate Retrieval (CAR) to evaluate the chronological understanding of the models. We decompose textual descriptions into events, and prepare negative text samples by shuffling the order of events in compound action descriptions. We then design a simple task for motion-language models to retrieve the more likely text from the ground truth and its chronologically shuffled version. CAR reveals many cases where current motion-language models fail to distinguish the event chronology of human motion, despite their impressive performance in terms of conventional evaluation metrics. To achieve better temporal alignment between text and motion, we further propose to use these texts with shuffled sequence of events as negative samples during training to reinforce the motion-language models. We conduct experiments on text-motion retrieval and text-to-motion generation using the reinforced motion-language models, which demonstrate improved performance over conventional approaches, indicating the necessity to consider temporal elements in motion-language alignment.

Learning to reliably perceive and understand the scene is an integral enabler for robots to operate in the real-world. This problem is inherently challenging due to the multitude of object types as well as appearance changes caused by varying illumination and weather conditions. Leveraging complementary modalities can enable learning of semantically richer representations that are resilient to such perturbations. Despite the tremendous progress in recent years, most multimodal convolutional neural network approaches directly concatenate feature maps from individual modality streams rendering the model incapable of focusing only on relevant complementary information for fusion. To address this limitation, we propose a mutimodal semantic segmentation framework that dynamically adapts the fusion of modality-specific features while being sensitive to the object category, spatial location and scene context in a self-supervised manner. Specifically, we propose an architecture consisting of two modality-specific encoder streams that fuse intermediate encoder representations into a single decoder using our proposed self-supervised model adaptation fusion mechanism which optimally combines complementary features. As intermediate representations are not aligned across modalities, we introduce an attention scheme for better correlation. In addition, we propose a computationally efficient unimodal segmentation architecture termed AdapNet++ that incorporates a new encoder with multiscale residual units and an efficient atrous spatial pyramid pooling that has a larger effective receptive field with more than 10x fewer parameters, complemented with a strong decoder with a multi-resolution supervision scheme that recovers high-resolution details. Comprehensive empirical evaluations on several benchmarks demonstrate that both our unimodal and multimodal architectures achieve state-of-the-art performance.

Early action prediction deals with inferring the ongoing action from partially-observed videos, typically at the outset of the video. We propose a bottleneck-based attention model that captures the evolution of the action, through progressive sampling over fine-to-coarse scales. Our proposed Temporal Progressive (TemPr) model is composed of multiple attention towers, one for each scale. The predicted action label is based on the collective agreement considering confidences of these towers. Extensive experiments over four video datasets showcase state-of-the-art performance on the task of Early Action Prediction across a range of encoder architectures. We demonstrate the effectiveness and consistency of TemPr through detailed ablations.

Recent vision-language-action (VLA) models rely on 2D inputs, lacking integration with the broader realm of the 3D physical world. Furthermore, they perform action prediction by learning a direct mapping from perception to action, neglecting the vast dynamics of the world and the relations between actions and dynamics. In contrast, human beings are endowed with world models that depict imagination about future scenarios to plan actions accordingly. To this end, we propose 3D-VLA by introducing a new family of embodied foundation models that seamlessly link 3D perception, reasoning, and action through a generative world model. Specifically, 3D-VLA is built on top of a 3D-based large language model (LLM), and a set of interaction tokens is introduced to engage with the embodied environment. Furthermore, to inject generation abilities into the model, we train a series of embodied diffusion models and align them into the LLM for predicting the goal images and point clouds. To train our 3D-VLA, we curate a large-scale 3D embodied instruction dataset by extracting vast 3D-related information from existing robotics datasets. Our experiments on held-in datasets demonstrate that 3D-VLA significantly improves the reasoning, multimodal generation, and planning capabilities in embodied environments, showcasing its potential in real-world applications.

Standard multi-modal models assume the use of the same modalities in training and inference stages. However, in practice, the environment in which multi-modal models operate may not satisfy such assumption. As such, their performances degrade drastically if any modality is missing in the inference stage. We ask: how can we train a model that is robust to missing modalities? This paper seeks a set of good practices for multi-modal action recognition, with a particular interest in circumstances where some modalities are not available at an inference time. First, we study how to effectively regularize the model during training (e.g., data augmentation). Second, we investigate on fusion methods for robustness to missing modalities: we find that transformer-based fusion shows better robustness for missing modality than summation or concatenation. Third, we propose a simple modular network, ActionMAE, which learns missing modality predictive coding by randomly dropping modality features and tries to reconstruct them with the remaining modality features. Coupling these good practices, we build a model that is not only effective in multi-modal action recognition but also robust to modality missing. Our model achieves the state-of-the-arts on multiple benchmarks and maintains competitive performances even in missing modality scenarios. Codes are available at https://github.com/sangminwoo/ActionMAE.

This study focuses on a novel task in text-to-image (T2I) generation, namely action customization. The objective of this task is to learn the co-existing action from limited data and generalize it to unseen humans or even animals. Experimental results show that existing subject-driven customization methods fail to learn the representative characteristics of actions and struggle in decoupling actions from context features, including appearance. To overcome the preference for low-level features and the entanglement of high-level features, we propose an inversion-based method Action-Disentangled Identifier (ADI) to learn action-specific identifiers from the exemplar images. ADI first expands the semantic conditioning space by introducing layer-wise identifier tokens, thereby increasing the representational richness while distributing the inversion across different features. Then, to block the inversion of action-agnostic features, ADI extracts the gradient invariance from the constructed sample triples and masks the updates of irrelevant channels. To comprehensively evaluate the task, we present an ActionBench that includes a variety of actions, each accompanied by meticulously selected samples. Both quantitative and qualitative results show that our ADI outperforms existing baselines in action-customized T2I generation. Our project page is at https://adi-t2i.github.io/ADI.

In the rapidly evolving landscape of AI research and application, Multimodal Large Language Models (MLLMs) have emerged as a transformative force, adept at interpreting and integrating information from diverse modalities such as text, images, and Graphical User Interfaces (GUIs). Despite these advancements, the nuanced interaction and understanding of GUIs pose a significant challenge, limiting the potential of existing models to enhance automation levels. To bridge this gap, this paper presents V-Zen, an innovative Multimodal Large Language Model (MLLM) meticulously crafted to revolutionise the domain of GUI understanding and grounding. Equipped with dual-resolution image encoders, V-Zen establishes new benchmarks in efficient grounding and next-action prediction, thereby laying the groundwork for self-operating computer systems. Complementing V-Zen is the GUIDE dataset, an extensive collection of real-world GUI elements and task-based sequences, serving as a catalyst for specialised fine-tuning. The successful integration of V-Zen and GUIDE marks the dawn of a new era in multimodal AI research, opening the door to intelligent, autonomous computing experiences. This paper extends an invitation to the research community to join this exciting journey, shaping the future of GUI automation. In the spirit of open science, our code, data, and model will be made publicly available, paving the way for multimodal dialogue scenarios with intricate and precise interactions.

Retrieving adverbs that describe an action in a video poses a crucial step towards fine-grained video understanding. We propose a framework for video-to-adverb retrieval (and vice versa) that aligns video embeddings with their matching compositional adverb-action text embedding in a joint embedding space. The compositional adverb-action text embedding is learned using a residual gating mechanism, along with a novel training objective consisting of triplet losses and a regression target. Our method achieves state-of-the-art performance on five recent benchmarks for video-adverb retrieval. Furthermore, we introduce dataset splits to benchmark video-adverb retrieval for unseen adverb-action compositions on subsets of the MSR-VTT Adverbs and ActivityNet Adverbs datasets. Our proposed framework outperforms all prior works for the generalisation task of retrieving adverbs from videos for unseen adverb-action compositions. Code and dataset splits are available at https://hummelth.github.io/ReGaDa/.

We present an unsupervised representation learning approach that compactly encodes the motion dependencies in videos. Given a pair of images from a video clip, our framework learns to predict the long-term 3D motions. To reduce the complexity of the learning framework, we propose to describe the motion as a sequence of atomic 3D flows computed with RGB-D modality. We use a Recurrent Neural Network based Encoder-Decoder framework to predict these sequences of flows. We argue that in order for the decoder to reconstruct these sequences, the encoder must learn a robust video representation that captures long-term motion dependencies and spatial-temporal relations. We demonstrate the effectiveness of our learned temporal representations on activity classification across multiple modalities and datasets such as NTU RGB+D and MSR Daily Activity 3D. Our framework is generic to any input modality, i.e., RGB, Depth, and RGB-D videos.

We investigate the internal representations of vision-and-language models (VLMs) and how they encode task representations. We consider tasks specified through examples or instructions, using either text or image inputs. Surprisingly, we find that conceptually similar tasks are mapped to similar task vector representations, regardless of how they are specified. Our findings suggest that to output answers, tokens in VLMs undergo three distinct phases: input, task, and answer, a process which is consistent across different modalities and specifications. The task vectors we identify in VLMs are general enough to be derived in one modality (e.g., text) and transferred to another (e.g., image). Additionally, we find that ensembling exemplar and instruction based task vectors produce better task representations. Taken together, these insights shed light on the underlying mechanisms of VLMs, particularly their ability to represent tasks in a shared manner across different modalities and task specifications. Project page: https://task-vectors-are-cross-modal.github.io.

This paper presents M^3GPT, an advanced Multimodal, Multitask framework for Motion comprehension and generation. M^3GPT operates on three fundamental principles. The first focuses on creating a unified representation space for various motion-relevant modalities. We employ discrete vector quantization for multimodal control and generation signals, such as text, music and motion/dance, enabling seamless integration into a large language model (LLM) with a single vocabulary. The second involves modeling model generation directly in the raw motion space. This strategy circumvents the information loss associated with discrete tokenizer, resulting in more detailed and comprehensive model generation. Third, M^3GPT learns to model the connections and synergies among various motion-relevant tasks. Text, the most familiar and well-understood modality for LLMs, is utilized as a bridge to establish connections between different motion tasks, facilitating mutual reinforcement. To our knowledge, M^3GPT is the first model capable of comprehending and generating motions based on multiple signals. Extensive experiments highlight M^3GPT's superior performance across various motion-relevant tasks and its powerful zero-shot generalization capabilities for extremely challenging tasks.

We introduce Latent Action Pretraining for general Action models (LAPA), an unsupervised method for pretraining Vision-Language-Action (VLA) models without ground-truth robot action labels. Existing Vision-Language-Action models require action labels typically collected by human teleoperators during pretraining, which significantly limits possible data sources and scale. In this work, we propose a method to learn from internet-scale videos that do not have robot action labels. We first train an action quantization model leveraging VQ-VAE-based objective to learn discrete latent actions between image frames, then pretrain a latent VLA model to predict these latent actions from observations and task descriptions, and finally finetune the VLA on small-scale robot manipulation data to map from latent to robot actions. Experimental results demonstrate that our method significantly outperforms existing techniques that train robot manipulation policies from large-scale videos. Furthermore, it outperforms the state-of-the-art VLA model trained with robotic action labels on real-world manipulation tasks that require language conditioning, generalization to unseen objects, and semantic generalization to unseen instructions. Training only on human manipulation videos also shows positive transfer, opening up the potential for leveraging web-scale data for robotics foundation model.

Weakly-supervised action localization aims to recognize and localize action instancese in untrimmed videos with only video-level labels. Most existing models rely on multiple instance learning(MIL), where the predictions of unlabeled instances are supervised by classifying labeled bags. The MIL-based methods are relatively well studied with cogent performance achieved on classification but not on localization. Generally, they locate temporal regions by the video-level classification but overlook the temporal variations of feature semantics. To address this problem, we propose a novel attention-based hierarchically-structured latent model to learn the temporal variations of feature semantics. Specifically, our model entails two components, the first is an unsupervised change-points detection module that detects change-points by learning the latent representations of video features in a temporal hierarchy based on their rates of change, and the second is an attention-based classification model that selects the change-points of the foreground as the boundaries. To evaluate the effectiveness of our model, we conduct extensive experiments on two benchmark datasets, THUMOS-14 and ActivityNet-v1.3. The experiments show that our method outperforms current state-of-the-art methods, and even achieves comparable performance with fully-supervised methods.

Diverse actions give rise to rich audio-visual signals in long videos. Recent works showcase that the two modalities of audio and video exhibit different temporal extents of events and distinct labels. We address the interplay between the two modalities in long videos by explicitly modelling the temporal extents of audio and visual events. We propose the Time Interval Machine (TIM) where a modality-specific time interval poses as a query to a transformer encoder that ingests a long video input. The encoder then attends to the specified interval, as well as the surrounding context in both modalities, in order to recognise the ongoing action. We test TIM on three long audio-visual video datasets: EPIC-KITCHENS, Perception Test, and AVE, reporting state-of-the-art (SOTA) for recognition. On EPIC-KITCHENS, we beat previous SOTA that utilises LLMs and significantly larger pre-training by 2.9% top-1 action recognition accuracy. Additionally, we show that TIM can be adapted for action detection, using dense multi-scale interval queries, outperforming SOTA on EPIC-KITCHENS-100 for most metrics, and showing strong performance on the Perception Test. Our ablations show the critical role of integrating the two modalities and modelling their time intervals in achieving this performance. Code and models at: https://github.com/JacobChalk/TIM

Active perception, a crucial human capability, involves setting a goal based on the current understanding of the environment and performing actions to achieve that goal. Despite significant efforts in evaluating Multimodal Large Language Models (MLLMs), active perception has been largely overlooked. To address this gap, we propose a novel benchmark named ActiView to evaluate active perception in MLLMs. Since comprehensively assessing active perception is challenging, we focus on a specialized form of Visual Question Answering (VQA) that eases the evaluation yet challenging for existing MLLMs. Given an image, we restrict the perceptual field of a model, requiring it to actively zoom or shift its perceptual field based on reasoning to answer the question successfully. We conduct extensive evaluation over 27 models, including proprietary and open-source models, and observe that the ability to read and comprehend multiple images simultaneously plays a significant role in enabling active perception. Results reveal a significant gap in the active perception capability of MLLMs, indicating that this area deserves more attention. We hope that our benchmark could help develop methods for MLLMs to understand multimodal inputs in more natural and holistic ways.

Zero-shot action recognition (ZSAR) aims to learn an alignment model between videos and class descriptions of seen actions that is transferable to unseen actions. The text queries (class descriptions) used in existing ZSAR works, however, are often short action names that fail to capture the rich semantics in the videos, leading to misalignment. With the intuition that video content descriptions (e.g., video captions) can provide rich contextual information of visual concepts in videos, we propose to utilize human annotated video descriptions to enrich the semantics of the class descriptions of each action. However, all existing action video description datasets are limited in terms of the number of actions, the semantics of video descriptions, etc. To this end, we collect a large-scale action video descriptions dataset named ActionHub, which covers a total of 1,211 common actions and provides 3.6 million action video descriptions. With the proposed ActionHub dataset, we further propose a novel Cross-modality and Cross-action Modeling (CoCo) framework for ZSAR, which consists of a Dual Cross-modality Alignment module and a Cross-action Invariance Mining module. Specifically, the Dual Cross-modality Alignment module utilizes both action labels and video descriptions from ActionHub to obtain rich class semantic features for feature alignment. The Cross-action Invariance Mining module exploits a cycle-reconstruction process between the class semantic feature spaces of seen actions and unseen actions, aiming to guide the model to learn cross-action invariant representations. Extensive experimental results demonstrate that our CoCo framework significantly outperforms the state-of-the-art on three popular ZSAR benchmarks (i.e., Kinetics-ZSAR, UCF101 and HMDB51) under two different learning protocols in ZSAR. We will release our code, models, and the proposed ActionHub dataset.

The advent of Multimodal LLMs has significantly enhanced image OCR recognition capabilities, making GUI automation a viable reality for increasing efficiency in digital tasks. One fundamental aspect of developing a GUI automation system is understanding primitive GUI actions. This comprehension is crucial as it enables agents to learn from user demonstrations, an essential element of automation. To rigorously evaluate such capabilities, we developed a video captioning benchmark for GUI actions, comprising 4,189 diverse video captioning samples. This task presents unique challenges compared to natural scene video captioning: 1) GUI screenshots typically contain denser information than natural scenes, and 2) events within GUIs are subtler and occur more rapidly, requiring precise attention to the appropriate time span and spatial region for accurate understanding. To address these challenges, we introduce our GUI action dataset Act2Cap as well as a simple yet effective framework, GUI Narrator, for GUI video captioning that utilizes the cursor as a visual prompt to enhance the interpretation of high-resolution screenshots. Specifically, a cursor detector is trained on our dataset, and a multimodal LLM model with mechanisms for selecting keyframes and key regions generates the captions. Experimental results indicate that even for today's most advanced multimodal models, such as GPT-4o, the task remains highly challenging. Additionally, our evaluations show that our strategy effectively enhances model performance, whether integrated into the fine-tuning of open-source models or employed as a prompting strategy in closed-source models.

Current video-based Masked Autoencoders (MAEs) primarily focus on learning effective spatiotemporal representations from a visual perspective, which may lead the model to prioritize general spatial-temporal patterns but often overlook nuanced semantic attributes like specific interactions or sequences that define actions - such as action-specific features that align more closely with human cognition for space-time correspondence. This can limit the model's ability to capture the essence of certain actions that are contextually rich and continuous. Humans are capable of mapping visual concepts, object view invariance, and semantic attributes available in static instances to comprehend natural dynamic scenes or videos. Existing MAEs for videos and static images rely on separate datasets for videos and images, which may lack the rich semantic attributes necessary for fully understanding the learned concepts, especially when compared to using video and corresponding sampled frame images together. To this end, we propose CrossVideoMAE an end-to-end self-supervised cross-modal contrastive learning MAE that effectively learns both video-level and frame-level rich spatiotemporal representations and semantic attributes. Our method integrates mutual spatiotemporal information from videos with spatial information from sampled frames within a feature-invariant space, while encouraging invariance to augmentations within the video domain. This objective is achieved through jointly embedding features of visible tokens and combining feature correspondence within and across modalities, which is critical for acquiring rich, label-free guiding signals from both video and frame image modalities in a self-supervised manner. Extensive experiments demonstrate that our approach surpasses previous state-of-the-art methods and ablation studies validate the effectiveness of our approach.

Recently, the rise of large-scale vision-language pretrained models like CLIP, coupled with the technology of Parameter-Efficient FineTuning (PEFT), has captured substantial attraction in video action recognition. Nevertheless, prevailing approaches tend to prioritize strong supervised performance at the expense of compromising the models' generalization capabilities during transfer. In this paper, we introduce a novel Multimodal, Multi-task CLIP adapting framework named \name to address these challenges, preserving both high supervised performance and robust transferability. Firstly, to enhance the individual modality architectures, we introduce multimodal adapters to both the visual and text branches. Specifically, we design a novel visual TED-Adapter, that performs global Temporal Enhancement and local temporal Difference modeling to improve the temporal representation capabilities of the visual encoder. Moreover, we adopt text encoder adapters to strengthen the learning of semantic label information. Secondly, we design a multi-task decoder with a rich set of supervisory signals to adeptly satisfy the need for strong supervised performance and generalization within a multimodal framework. Experimental results validate the efficacy of our approach, demonstrating exceptional performance in supervised learning while maintaining strong generalization in zero-shot scenarios.

This study delves into the realm of multi-modality (i.e., video and motion modalities) human behavior understanding by leveraging the powerful capabilities of Large Language Models (LLMs). Diverging from recent LLMs designed for video-only or motion-only understanding, we argue that understanding human behavior necessitates joint modeling from both videos and motion sequences (e.g., SMPL sequences) to capture nuanced body part dynamics and semantics effectively. In light of this, we present MotionLLM, a straightforward yet effective framework for human motion understanding, captioning, and reasoning. Specifically, MotionLLM adopts a unified video-motion training strategy that leverages the complementary advantages of existing coarse video-text data and fine-grained motion-text data to glean rich spatial-temporal insights. Furthermore, we collect a substantial dataset, MoVid, comprising diverse videos, motions, captions, and instructions. Additionally, we propose the MoVid-Bench, with carefully manual annotations, for better evaluation of human behavior understanding on video and motion. Extensive experiments show the superiority of MotionLLM in the caption, spatial-temporal comprehension, and reasoning ability.

We introduce CLEVR-Math, a multi-modal math word problems dataset consisting of simple math word problems involving addition/subtraction, represented partly by a textual description and partly by an image illustrating the scenario. The text describes actions performed on the scene that is depicted in the image. Since the question posed may not be about the scene in the image, but about the state of the scene before or after the actions are applied, the solver envision or imagine the state changes due to these actions. Solving these word problems requires a combination of language, visual and mathematical reasoning. We apply state-of-the-art neural and neuro-symbolic models for visual question answering on CLEVR-Math and empirically evaluate their performances. Our results show how neither method generalise to chains of operations. We discuss the limitations of the two in addressing the task of multi-modal word problem solving.

Self-supervised learning (SSL) has revolutionized visual representation learning, but has not achieved the robustness of human vision. A reason for this could be that SSL does not leverage all the data available to humans during learning. When learning about an object, humans often purposefully turn or move around objects and research suggests that these interactions can substantially enhance their learning. Here we explore whether such object-related actions can boost SSL. For this, we extract the actions performed to change from one ego-centric view of an object to another in four video datasets. We then introduce a new loss function to learn visual and action embeddings by aligning the performed action with the representations of two images extracted from the same clip. This permits the performed actions to structure the latent visual representation. Our experiments show that our method consistently outperforms previous methods on downstream category recognition. In our analysis, we find that the observed improvement is associated with a better viewpoint-wise alignment of different objects from the same category. Overall, our work demonstrates that embodied interactions with objects can improve SSL of object categories.

The goal of building a benchmark (suite of datasets) is to provide a unified protocol for fair evaluation and thus facilitate the evolution of a specific area. Nonetheless, we point out that existing protocols of action recognition could yield partial evaluations due to several limitations. To comprehensively probe the effectiveness of spatiotemporal representation learning, we introduce BEAR, a new BEnchmark on video Action Recognition. BEAR is a collection of 18 video datasets grouped into 5 categories (anomaly, gesture, daily, sports, and instructional), which covers a diverse set of real-world applications. With BEAR, we thoroughly evaluate 6 common spatiotemporal models pre-trained by both supervised and self-supervised learning. We also report transfer performance via standard finetuning, few-shot finetuning, and unsupervised domain adaptation. Our observation suggests that current state-of-the-art cannot solidly guarantee high performance on datasets close to real-world applications, and we hope BEAR can serve as a fair and challenging evaluation benchmark to gain insights on building next-generation spatiotemporal learners. Our dataset, code, and models are released at: https://github.com/AndongDeng/BEAR

We introduce Cambrian-1, a family of multimodal LLMs (MLLMs) designed with a vision-centric approach. While stronger language models can enhance multimodal capabilities, the design choices for vision components are often insufficiently explored and disconnected from visual representation learning research. This gap hinders accurate sensory grounding in real-world scenarios. Our study uses LLMs and visual instruction tuning as an interface to evaluate various visual representations, offering new insights into different models and architectures -- self-supervised, strongly supervised, or combinations thereof -- based on experiments with over 20 vision encoders. We critically examine existing MLLM benchmarks, addressing the difficulties involved in consolidating and interpreting results from various tasks, and introduce a new vision-centric benchmark, CV-Bench. To further improve visual grounding, we propose the Spatial Vision Aggregator (SVA), a dynamic and spatially-aware connector that integrates high-resolution vision features with LLMs while reducing the number of tokens. Additionally, we discuss the curation of high-quality visual instruction-tuning data from publicly available sources, emphasizing the importance of data source balancing and distribution ratio. Collectively, Cambrian-1 not only achieves state-of-the-art performance but also serves as a comprehensive, open cookbook for instruction-tuned MLLMs. We provide model weights, code, supporting tools, datasets, and detailed instruction-tuning and evaluation recipes. We hope our release will inspire and accelerate advancements in multimodal systems and visual representation learning.

We propose MM-REACT, a system paradigm that integrates ChatGPT with a pool of vision experts to achieve multimodal reasoning and action. In this paper, we define and explore a comprehensive list of advanced vision tasks that are intriguing to solve, but may exceed the capabilities of existing vision and vision-language models. To achieve such advanced visual intelligence, MM-REACT introduces a textual prompt design that can represent text descriptions, textualized spatial coordinates, and aligned file names for dense visual signals such as images and videos. MM-REACT's prompt design allows language models to accept, associate, and process multimodal information, thereby facilitating the synergetic combination of ChatGPT and various vision experts. Zero-shot experiments demonstrate MM-REACT's effectiveness in addressing the specified capabilities of interests and its wide application in different scenarios that require advanced visual understanding. Furthermore, we discuss and compare MM-REACT's system paradigm with an alternative approach that extends language models for multimodal scenarios through joint finetuning. Code, demo, video, and visualization are available at https://multimodal-react.github.io/

We present Unified-IO 2, the first autoregressive multimodal model that is capable of understanding and generating image, text, audio, and action. To unify different modalities, we tokenize inputs and outputs -- images, text, audio, action, bounding boxes, etc., into a shared semantic space and then process them with a single encoder-decoder transformer model. Since training with such diverse modalities is challenging, we propose various architectural improvements to stabilize model training. We train our model from scratch on a large multimodal pre-training corpus from diverse sources with a multimodal mixture of denoisers objective. To learn an expansive set of skills, such as following multimodal instructions, we construct and finetune on an ensemble of 120 datasets with prompts and augmentations. With a single unified model, Unified-IO 2 achieves state-of-the-art performance on the GRIT benchmark and strong results in more than 35 benchmarks, including image generation and understanding, natural language understanding, video and audio understanding, and robotic manipulation. We release all our models to the research community.

The recent success of interleaved Large Multimodal Models (LMMs) in few-shot learning suggests that in-context learning (ICL) with many examples can be promising for learning new tasks. However, this many-shot multimodal ICL setting has one crucial problem: it is fundamentally limited by the model's context length set at pretraining. The problem is especially prominent in the multimodal domain, which processes both text and images, requiring additional tokens. This motivates the need for a multimodal method to compress many shots into fewer tokens without finetuning. In this work, we enable LMMs to perform multimodal, many-shot in-context learning by leveraging Multimodal Task Vectors (MTV)--compact implicit representations of in-context examples compressed in the model's attention heads. Specifically, we first demonstrate the existence of such MTV in LMMs and then leverage these extracted MTV to enable many-shot in-context learning for various vision-and-language tasks. Our experiments suggest that MTV can scale in performance with the number of compressed shots and generalize to similar out-of-domain tasks without additional context length for inference.

The goal of spatial-temporal action detection is to determine the time and place where each person's action occurs in a video and classify the corresponding action category. Most of the existing methods adopt fully-supervised learning, which requires a large amount of training data, making it very difficult to achieve zero-shot learning. In this paper, we propose to utilize a pre-trained visual-language model to extract the representative image and text features, and model the relationship between these features through different interaction modules to obtain the interaction feature. In addition, we use this feature to prompt each label to obtain more appropriate text features. Finally, we calculate the similarity between the interaction feature and the text feature for each label to determine the action category. Our experiments on J-HMDB and UCF101-24 datasets demonstrate that the proposed interaction module and prompting make the visual-language features better aligned, thus achieving excellent accuracy for zero-shot spatio-temporal action detection. The code will be available at https://github.com/webber2933/iCLIP.

Modeling long-term context in videos is crucial for many fine-grained tasks including temporal action segmentation. An interesting question that is still open is how much long-term temporal context is needed for optimal performance. While transformers can model the long-term context of a video, this becomes computationally prohibitive for long videos. Recent works on temporal action segmentation thus combine temporal convolutional networks with self-attentions that are computed only for a local temporal window. While these approaches show good results, their performance is limited by their inability to capture the full context of a video. In this work, we try to answer how much long-term temporal context is required for temporal action segmentation by introducing a transformer-based model that leverages sparse attention to capture the full context of a video. We compare our model with the current state of the art on three datasets for temporal action segmentation, namely 50Salads, Breakfast, and Assembly101. Our experiments show that modeling the full context of a video is necessary to obtain the best performance for temporal action segmentation.

Large language model (LLM) leads to a surge of autonomous GUI agents for smartphone, which completes a task triggered by natural language through predicting a sequence of actions of API. Even though the task highly relies on past actions and visual observations, existing studies typical consider little semantic information carried out by intermediate screenshots and screen operations. To address this, this work presents Chain-of-Action-Thought (dubbed CoAT), which takes the description of the previous actions, the current screen, and more importantly the action thinking of what actions should be performed and the outcomes led by the chosen action. We demonstrate that, in a zero-shot setting upon an off-the-shell LLM, CoAT significantly improves the goal progress compared to standard context modeling. To further facilitate the research in this line, we construct a benchmark Android-In-The-Zoo (AitZ), which contains 18,643 screen-action pairs together with chain-of-action-thought annotations. Experiments show that fine-tuning a 200M model on our AitZ dataset achieves on par performance with CogAgent-Chat-18B.

Generative Large Multimodal Models (LMMs) like LLaVA and Qwen-VL excel at a wide variety of vision-language (VL) tasks such as image captioning or visual question answering. Despite strong performance, LMMs are not directly suited for foundational discriminative vision-language tasks (i.e., tasks requiring discrete label predictions) such as image classification and multiple-choice VQA. One key challenge in utilizing LMMs for discriminative tasks is the extraction of useful features from generative models. To overcome this issue, we propose an approach for finding features in the model's latent space to more effectively leverage LMMs for discriminative tasks. Toward this end, we present Sparse Attention Vectors (SAVs) -- a finetuning-free method that leverages sparse attention head activations (fewer than 1\% of the heads) in LMMs as strong features for VL tasks. With only few-shot examples, SAVs demonstrate state-of-the-art performance compared to a variety of few-shot and finetuned baselines on a collection of discriminative tasks. Our experiments also imply that SAVs can scale in performance with additional examples and generalize to similar tasks, establishing SAVs as both effective and robust multimodal feature representations.

Large multimodal models have demonstrated impressive problem-solving abilities in vision and language tasks, and have the potential to encode extensive world knowledge. However, it remains an open challenge for these models to perceive, reason, plan, and act in realistic environments. In this work, we introduce Can-Do, a benchmark dataset designed to evaluate embodied planning abilities through more diverse and complex scenarios than previous datasets. Our dataset includes 400 multimodal samples, each consisting of natural language user instructions, visual images depicting the environment, state changes, and corresponding action plans. The data encompasses diverse aspects of commonsense knowledge, physical understanding, and safety awareness. Our fine-grained analysis reveals that state-of-the-art models, including GPT-4V, face bottlenecks in visual perception, comprehension, and reasoning abilities. To address these challenges, we propose NeuroGround, a neurosymbolic framework that first grounds the plan generation in the perceived environment states and then leverages symbolic planning engines to augment the model-generated plans. Experimental results demonstrate the effectiveness of our framework compared to strong baselines. Our code and dataset are available at https://embodied-planning.github.io.

Understanding human behavior requires measuring behavioral actions. Due to its complexity, behavior is best mapped onto a rich, semantic structure such as language. The recent development of multi-modal large language models (MLLMs) is a promising candidate for a wide range of action understanding tasks. In this work, we focus on evaluating and then improving MLLMs to perform action recognition. We reformulate EPIC-KITCHENS-100, one of the largest and most challenging egocentric action datasets, to the form of video multiple question answering (EPIC-KITCHENS-100-MQA). We show that when we sample difficult incorrect answers as distractors, leading MLLMs struggle to recognize the correct actions. We propose a series of methods that greatly improve the MLLMs' ability to perform action recognition, achieving state-of-the-art on both the EPIC-KITCHENS-100 validation set, as well as outperforming GPT-4o by 21 points in accuracy on EPIC-KITCHENS-100-MQA. Lastly, we show improvements on other action-related video benchmarks such as EgoSchema, PerceptionTest, LongVideoBench, VideoMME and MVBench, suggesting that MLLMs are a promising path forward for complex action tasks. Code and models are available at: https://github.com/AdaptiveMotorControlLab/LLaVAction.

Pre-training on large-scale video data has become a common recipe for learning transferable spatiotemporal representations in recent years. Despite some progress, existing methods are mostly limited to highly curated datasets (e.g., K400) and exhibit unsatisfactory out-of-the-box representations. We argue that it is due to the fact that they only capture pixel-level knowledge rather than spatiotemporal semantics, which hinders further progress in video understanding. Inspired by the great success of image-text pre-training (e.g., CLIP), we take the first step to exploit language semantics to boost transferable spatiotemporal representation learning. We introduce a new pretext task, Turning to Video for Transcript Sorting (TVTS), which sorts shuffled ASR scripts by attending to learned video representations. We do not rely on descriptive captions and learn purely from video, i.e., leveraging the natural transcribed speech knowledge to provide noisy but useful semantics over time. Our method enforces the vision model to contextualize what is happening over time so that it can re-organize the narrative transcripts, and can seamlessly apply to large-scale uncurated video data in the real world. Our method demonstrates strong out-of-the-box spatiotemporal representations on diverse benchmarks, e.g., +13.6% gains over VideoMAE on SSV2 via linear probing. The code is available at https://github.com/TencentARC/TVTS.

Large language models excel at a wide range of complex tasks. However, enabling general inference in the real world, e.g., for robotics problems, raises the challenge of grounding. We propose embodied language models to directly incorporate real-world continuous sensor modalities into language models and thereby establish the link between words and percepts. Input to our embodied language model are multi-modal sentences that interleave visual, continuous state estimation, and textual input encodings. We train these encodings end-to-end, in conjunction with a pre-trained large language model, for multiple embodied tasks including sequential robotic manipulation planning, visual question answering, and captioning. Our evaluations show that PaLM-E, a single large embodied multimodal model, can address a variety of embodied reasoning tasks, from a variety of observation modalities, on multiple embodiments, and further, exhibits positive transfer: the model benefits from diverse joint training across internet-scale language, vision, and visual-language domains. Our largest model, PaLM-E-562B with 562B parameters, in addition to being trained on robotics tasks, is a visual-language generalist with state-of-the-art performance on OK-VQA, and retains generalist language capabilities with increasing scale.

In recent years, instruction-tuned Large Multimodal Models (LMMs) have been successful at several tasks, including image captioning and visual question answering; yet leveraging these models remains an open question for robotics. Prior LMMs for robotics applications have been extensively trained on language and action data, but their ability to generalize in different settings has often been less than desired. To address this, we introduce LLARVA, a model trained with a novel instruction tuning method that leverages structured prompts to unify a range of robotic learning tasks, scenarios, and environments. Additionally, we show that predicting intermediate 2-D representations, which we refer to as "visual traces", can help further align vision and action spaces for robot learning. We generate 8.5M image-visual trace pairs from the Open X-Embodiment dataset in order to pre-train our model, and we evaluate on 12 different tasks in the RLBench simulator as well as a physical Franka Emika Panda 7-DoF robot. Our experiments yield strong performance, demonstrating that LLARVA - using 2-D and language representations - performs well compared to several contemporary baselines, and can generalize across various robot environments and configurations.

Large Multimodal Models (LMMs) excel in natural language and visual understanding but are challenged by exacting tasks such as Knowledge-based Visual Question Answering (KB-VQA) which involve the retrieval of relevant information from document collections to use in shaping answers to questions. We present an extensive training and evaluation framework, M2KR, for KB-VQA. M2KR contains a collection of vision and language tasks which we have incorporated into a single suite of benchmark tasks for training and evaluating general-purpose multi-modal retrievers. We use M2KR to develop PreFLMR, a pre-trained version of the recently developed Fine-grained Late-interaction Multi-modal Retriever (FLMR) approach to KB-VQA, and we report new state-of-the-art results across a range of tasks. We also present investigations into the scaling behaviors of PreFLMR intended to be useful in future developments in general-purpose multi-modal retrievers.

While existing large vision-language multimodal models focus on whole image understanding, there is a prominent gap in achieving region-specific comprehension. Current approaches that use textual coordinates or spatial encodings often fail to provide a user-friendly interface for visual prompting. To address this challenge, we introduce a novel multimodal model capable of decoding arbitrary visual prompts. This allows users to intuitively mark images and interact with the model using natural cues like a "red bounding box" or "pointed arrow". Our simple design directly overlays visual markers onto the RGB image, eliminating the need for complex region encodings, yet achieves state-of-the-art performance on region-understanding tasks like Visual7W, PointQA, and Visual Commonsense Reasoning benchmark. Furthermore, we present ViP-Bench, a comprehensive benchmark to assess the capability of models in understanding visual prompts across multiple dimensions, enabling future research in this domain. Code, data, and model are publicly available.

One significant factor we expect the video representation learning to capture, especially in contrast with the image representation learning, is the object motion. However, we found that in the current mainstream video datasets, some action categories are highly related with the scene where the action happens, making the model tend to degrade to a solution where only the scene information is encoded. For example, a trained model may predict a video as playing football simply because it sees the field, neglecting that the subject is dancing as a cheerleader on the field. This is against our original intention towards the video representation learning and may bring scene bias on different dataset that can not be ignored. In order to tackle this problem, we propose to decouple the scene and the motion (DSM) with two simple operations, so that the model attention towards the motion information is better paid. Specifically, we construct a positive clip and a negative clip for each video. Compared to the original video, the positive/negative is motion-untouched/broken but scene-broken/untouched by Spatial Local Disturbance and Temporal Local Disturbance. Our objective is to pull the positive closer while pushing the negative farther to the original clip in the latent space. In this way, the impact of the scene is weakened while the temporal sensitivity of the network is further enhanced. We conduct experiments on two tasks with various backbones and different pre-training datasets, and find that our method surpass the SOTA methods with a remarkable 8.1% and 8.8% improvement towards action recognition task on the UCF101 and HMDB51 datasets respectively using the same backbone.

Understanding what sequence of steps are needed to complete a goal can help artificial intelligence systems reason about human activities. Past work in NLP has examined the task of goal-step inference for text. We introduce the visual analogue. We propose the Visual Goal-Step Inference (VGSI) task, where a model is given a textual goal and must choose which of four images represents a plausible step towards that goal. With a new dataset harvested from wikiHow consisting of 772,277 images representing human actions, we show that our task is challenging for state-of-the-art multimodal models. Moreover, the multimodal representation learned from our data can be effectively transferred to other datasets like HowTo100m, increasing the VGSI accuracy by 15 - 20%. Our task will facilitate multimodal reasoning about procedural events.

Large multimodal models (LMMs) combine unimodal encoders and large language models (LLMs) to perform multimodal tasks. Despite recent advancements towards the interpretability of these models, understanding internal representations of LMMs remains largely a mystery. In this paper, we present a novel framework for the interpretation of LMMs. We propose a dictionary learning based approach, applied to the representation of tokens. The elements of the learned dictionary correspond to our proposed concepts. We show that these concepts are well semantically grounded in both vision and text. Thus we refer to these as ``multi-modal concepts''. We qualitatively and quantitatively evaluate the results of the learnt concepts. We show that the extracted multimodal concepts are useful to interpret representations of test samples. Finally, we evaluate the disentanglement between different concepts and the quality of grounding concepts visually and textually. Our code is publicly available at https://github.com/mshukor/xl-vlms

This paper presents a novel approach to Zero-Shot Action Recognition. Recent works have explored the detection and classification of objects to obtain semantic information from videos with remarkable performance. Inspired by them, we propose using video captioning methods to extract semantic information about objects, scenes, humans, and their relationships. To the best of our knowledge, this is the first work to represent both videos and labels with descriptive sentences. More specifically, we represent videos using sentences generated via video captioning methods and classes using sentences extracted from documents acquired through search engines on the Internet. Using these representations, we build a shared semantic space employing BERT-based embedders pre-trained in the paraphrasing task on multiple text datasets. The projection of both visual and semantic information onto this space is straightforward, as they are sentences, enabling classification using the nearest neighbor rule. We demonstrate that representing videos and labels with sentences alleviates the domain adaptation problem. Additionally, we show that word vectors are unsuitable for building the semantic embedding space of our descriptions. Our method outperforms the state-of-the-art performance on the UCF101 dataset by 3.3 p.p. in accuracy under the TruZe protocol and achieves competitive results on both the UCF101 and HMDB51 datasets under the conventional protocol (0/50\% - training/testing split). Our code is available at https://github.com/valterlej/zsarcap.

Deep models that are both effective and explainable are desirable in many settings; prior explainable models have been unimodal, offering either image-based visualization of attention weights or text-based generation of post-hoc justifications. We propose a multimodal approach to explanation, and argue that the two modalities provide complementary explanatory strengths. We collect two new datasets to define and evaluate this task, and propose a novel model which can provide joint textual rationale generation and attention visualization. Our datasets define visual and textual justifications of a classification decision for activity recognition tasks (ACT-X) and for visual question answering tasks (VQA-X). We quantitatively show that training with the textual explanations not only yields better textual justification models, but also better localizes the evidence that supports the decision. We also qualitatively show cases where visual explanation is more insightful than textual explanation, and vice versa, supporting our thesis that multimodal explanation models offer significant benefits over unimodal approaches.

Recent advancements in Multimodal Large Language Models (MLLMs) have demonstrated remarkable progress in visual understanding. This impressive leap raises a compelling question: how can language models, initially trained solely on linguistic data, effectively interpret and process visual content? This paper aims to address this question with systematic investigation across 4 model families and 4 model scales, uncovering a unique class of attention heads that focus specifically on visual content. Our analysis reveals a strong correlation between the behavior of these attention heads, the distribution of attention weights, and their concentration on visual tokens within the input. These findings enhance our understanding of how LLMs adapt to multimodal tasks, demonstrating their potential to bridge the gap between textual and visual understanding. This work paves the way for the development of AI systems capable of engaging with diverse modalities.

We present an unsupervised representation learning approach using videos without semantic labels. We leverage the temporal coherence as a supervisory signal by formulating representation learning as a sequence sorting task. We take temporally shuffled frames (i.e., in non-chronological order) as inputs and train a convolutional neural network to sort the shuffled sequences. Similar to comparison-based sorting algorithms, we propose to extract features from all frame pairs and aggregate them to predict the correct order. As sorting shuffled image sequence requires an understanding of the statistical temporal structure of images, training with such a proxy task allows us to learn rich and generalizable visual representation. We validate the effectiveness of the learned representation using our method as pre-training on high-level recognition problems. The experimental results show that our method compares favorably against state-of-the-art methods on action recognition, image classification and object detection tasks.

Recent advances in Large Multimodal Models (LMMs) lead to significant breakthroughs in both academia and industry. One question that arises is how we, as humans, can understand their internal neural representations. This paper takes an initial step towards addressing this question by presenting a versatile framework to identify and interpret the semantics within LMMs. Specifically, 1) we first apply a Sparse Autoencoder(SAE) to disentangle the representations into human understandable features. 2) We then present an automatic interpretation framework to interpreted the open-semantic features learned in SAE by the LMMs themselves. We employ this framework to analyze the LLaVA-NeXT-8B model using the LLaVA-OV-72B model, demonstrating that these features can effectively steer the model's behavior. Our results contribute to a deeper understanding of why LMMs excel in specific tasks, including EQ tests, and illuminate the nature of their mistakes along with potential strategies for their rectification. These findings offer new insights into the internal mechanisms of LMMs and suggest parallels with the cognitive processes of the human brain.

There is increasing interest in the use of multimodal data in various web applications, such as digital advertising and e-commerce. Typical methods for extracting important information from multimodal data rely on a mid-fusion architecture that combines the feature representations from multiple encoders. However, as the number of modalities increases, several potential problems with the mid-fusion model structure arise, such as an increase in the dimensionality of the concatenated multimodal features and missing modalities. To address these problems, we propose a new concept that considers multimodal inputs as a set of sequences, namely, deep multimodal sequence sets (DM^2S^2). Our set-aware concept consists of three components that capture the relationships among multiple modalities: (a) a BERT-based encoder to handle the inter- and intra-order of elements in the sequences, (b) intra-modality residual attention (IntraMRA) to capture the importance of the elements in a modality, and (c) inter-modality residual attention (InterMRA) to enhance the importance of elements with modality-level granularity further. Our concept exhibits performance that is comparable to or better than the previous set-aware models. Furthermore, we demonstrate that the visualization of the learned InterMRA and IntraMRA weights can provide an interpretation of the prediction results.

We present the "Law of Vision Representation" in multimodal large language models (MLLMs). It reveals a strong correlation between the combination of cross-modal alignment, correspondence in vision representation, and MLLM performance. We quantify the two factors using the cross-modal Alignment and Correspondence score (AC score). Through extensive experiments involving thirteen different vision representation settings and evaluations across eight benchmarks, we find that the AC score is linearly correlated to model performance. By leveraging this relationship, we are able to identify and train the optimal vision representation only, which does not require finetuning the language model every time, resulting in a 99.7% reduction in computational cost.

Existing text-to-video diffusion models rely solely on text-only encoders for their pretraining. This limitation stems from the absence of large-scale multimodal prompt video datasets, resulting in a lack of visual grounding and restricting their versatility and application in multimodal integration. To address this, we construct a large-scale multimodal prompt dataset by employing retrieval methods to pair in-context examples with the given text prompts and then utilize a two-stage training strategy to enable diverse video generation tasks within the same model. In the first stage, we propose a multimodal conditional video generation framework for pretraining on these augmented datasets, establishing a foundational model for grounded video generation. Secondly, we finetune the model from the first stage on three video generation tasks, incorporating multi-modal instructions. This process further refines the model's ability to handle diverse inputs and tasks, ensuring seamless integration of multi-modal information. After this two-stage train-ing process, VIMI demonstrates multimodal understanding capabilities, producing contextually rich and personalized videos grounded in the provided inputs, as shown in Figure 1. Compared to previous visual grounded video generation methods, VIMI can synthesize consistent and temporally coherent videos with large motion while retaining the semantic control. Lastly, VIMI also achieves state-of-the-art text-to-video generation results on UCF101 benchmark.

The field of visual representation learning has seen explosive growth in the past years, but its benefits in robotics have been surprisingly limited so far. Prior work uses generic visual representations as a basis to learn (task-specific) robot action policies (e.g., via behavior cloning). While the visual representations do accelerate learning, they are primarily used to encode visual observations. Thus, action information has to be derived purely from robot data, which is expensive to collect! In this work, we present a scalable alternative where the visual representations can help directly infer robot actions. We observe that vision encoders express relationships between image observations as distances (e.g., via embedding dot product) that could be used to efficiently plan robot behavior. We operationalize this insight and develop a simple algorithm for acquiring a distance function and dynamics predictor, by fine-tuning a pre-trained representation on human collected video sequences. The final method is able to substantially outperform traditional robot learning baselines (e.g., 70% success v.s. 50% for behavior cloning on pick-place) on a suite of diverse real-world manipulation tasks. It can also generalize to novel objects, without using any robot demonstrations during train time. For visualizations of the learned policies please check: https://agi-labs.github.io/manipulate-by-seeing/.

Efficiently retrieving and synthesizing information from large-scale multimodal collections has become a critical challenge. However, existing video retrieval datasets suffer from scope limitations, primarily focusing on matching descriptive but vague queries with small collections of professionally edited, English-centric videos. To address this gap, we introduce MultiVENT 2.0, a large-scale, multilingual event-centric video retrieval benchmark featuring a collection of more than 218,000 news videos and 3,906 queries targeting specific world events. These queries specifically target information found in the visual content, audio, embedded text, and text metadata of the videos, requiring systems leverage all these sources to succeed at the task. Preliminary results show that state-of-the-art vision-language models struggle significantly with this task, and while alternative approaches show promise, they are still insufficient to adequately address this problem. These findings underscore the need for more robust multimodal retrieval systems, as effective video retrieval is a crucial step towards multimodal content understanding and generation tasks.

Understanding multimodal perception for embodied AI is an open question because such inputs may contain highly complementary as well as redundant information for the task. A relevant direction for multimodal policies is understanding the global trends of each modality at the fusion layer. To this end, we disentangle the attributions for visual, language, and previous action inputs across different policies trained on the ALFRED dataset. Attribution analysis can be utilized to rank and group the failure scenarios, investigate modeling and dataset biases, and critically analyze multimodal EAI policies for robustness and user trust before deployment. We present MAEA, a framework to compute global attributions per modality of any differentiable policy. In addition, we show how attributions enable lower-level behavior analysis in EAI policies for language and visual attributions.

Existing video-language pre-training methods primarily focus on instance-level alignment between video clips and captions via global contrastive learning but neglect rich fine-grained local information in both videos and text, which is of importance to downstream tasks requiring temporal localization and semantic reasoning. A powerful model is expected to be capable of capturing region-object correspondences and recognizing scene changes in a video clip, reflecting spatial and temporal granularity, respectively. To strengthen model's understanding into such fine-grained details, we propose a simple yet effective video-language modeling framework, S-ViLM, by exploiting the intrinsic structures of these two modalities. It includes two novel designs, inter-clip spatial grounding and intra-clip temporal grouping, to promote learning region-object alignment and temporal-aware features, simultaneously. Comprehensive evaluations demonstrate that S-ViLM performs favorably against existing approaches in learning more expressive representations. Specifically, S-ViLM surpasses the state-of-the-art methods substantially on four representative downstream tasks, covering text-video retrieval, video question answering, video action recognition, and temporal action localization.

We study unsupervised video representation learning that seeks to learn both motion and appearance features from unlabeled video only, which can be reused for downstream tasks such as action recognition. This task, however, is extremely challenging due to 1) the highly complex spatial-temporal information in videos; and 2) the lack of labeled data for training. Unlike the representation learning for static images, it is difficult to construct a suitable self-supervised task to well model both motion and appearance features. More recently, several attempts have been made to learn video representation through video playback speed prediction. However, it is non-trivial to obtain precise speed labels for the videos. More critically, the learnt models may tend to focus on motion pattern and thus may not learn appearance features well. In this paper, we observe that the relative playback speed is more consistent with motion pattern, and thus provide more effective and stable supervision for representation learning. Therefore, we propose a new way to perceive the playback speed and exploit the relative speed between two video clips as labels. In this way, we are able to well perceive speed and learn better motion features. Moreover, to ensure the learning of appearance features, we further propose an appearance-focused task, where we enforce the model to perceive the appearance difference between two video clips. We show that optimizing the two tasks jointly consistently improves the performance on two downstream tasks, namely action recognition and video retrieval. Remarkably, for action recognition on UCF101 dataset, we achieve 93.7% accuracy without the use of labeled data for pre-training, which outperforms the ImageNet supervised pre-trained model. Code and pre-trained models can be found at https://github.com/PeihaoChen/RSPNet.

Contrastive learning has shown promising potential for learning robust representations by utilizing unlabeled data. However, constructing effective positive-negative pairs for contrastive learning on facial behavior datasets remains challenging. This is because such pairs inevitably encode the subject-ID information, and the randomly constructed pairs may push similar facial images away due to the limited number of subjects in facial behavior datasets. To address this issue, we propose to utilize activity descriptions, coarse-grained information provided in some datasets, which can provide high-level semantic information about the image sequences but is often neglected in previous studies. More specifically, we introduce a two-stage Contrastive Learning with Text-Embeded framework for Facial behavior understanding (CLEF). The first stage is a weakly-supervised contrastive learning method that learns representations from positive-negative pairs constructed using coarse-grained activity information. The second stage aims to train the recognition of facial expressions or facial action units by maximizing the similarity between image and the corresponding text label names. The proposed CLEF achieves state-of-the-art performance on three in-the-lab datasets for AU recognition and three in-the-wild datasets for facial expression recognition.

Multimodal models have gained significant success in recent years. Standard multimodal approaches often assume unchanged modalities from training stage to inference stage. In practice, however, many scenarios fail to satisfy such assumptions with missing modalities during inference, leading to limitations on where multimodal models can be applied. While existing methods mitigate the problem through reconstructing the missing modalities, it increases unnecessary computational cost, which could be just as critical, especially for large, deployed systems. To solve the problem from both sides, we propose a novel guidance network that promotes knowledge sharing during training, taking advantage of the multimodal representations to train better single-modality models for inference. Real-life experiment in violence detection shows that our proposed framework trains single-modality models that significantly outperform its traditionally trained counterparts while maintaining the same inference cost.

Foundation Vision Language Models (VLMs) exhibit strong capabilities in multi-modal representation learning, comprehension, and reasoning. By injecting action components into the VLMs, Vision-Language-Action Models (VLAs) can be naturally formed and also show promising performance. Existing work has demonstrated the effectiveness and generalization of VLAs in multiple scenarios and tasks. Nevertheless, the transfer from VLMs to VLAs is not trivial since existing VLAs differ in their backbones, action-prediction formulations, data distributions, and training recipes. This leads to a missing piece for a systematic understanding of the design choices of VLAs. In this work, we disclose the key factors that significantly influence the performance of VLA and focus on answering three essential design choices: which backbone to select, how to formulate the VLA architectures, and when to add cross-embodiment data. The obtained results convince us firmly to explain why we need VLA and develop a new family of VLAs, RoboVLMs, which require very few manual designs and achieve a new state-of-the-art performance in three simulation tasks and real-world experiments. Through our extensive experiments, which include over 8 VLM backbones, 4 policy architectures, and over 600 distinct designed experiments, we provide a detailed guidebook for the future design of VLAs. In addition to the study, the highly flexible RoboVLMs framework, which supports easy integrations of new VLMs and free combinations of various design choices, is made public to facilitate future research. We open-source all details, including codes, models, datasets, and toolkits, along with detailed training and evaluation recipes at: robovlms.github.io.

Though the advancement of pre-trained large language models unfolds, the exploration of building a unified model for language and other multi-modal data, such as motion, remains challenging and untouched so far. Fortunately, human motion displays a semantic coupling akin to human language, often perceived as a form of body language. By fusing language data with large-scale motion models, motion-language pre-training that can enhance the performance of motion-related tasks becomes feasible. Driven by this insight, we propose MotionGPT, a unified, versatile, and user-friendly motion-language model to handle multiple motion-relevant tasks. Specifically, we employ the discrete vector quantization for human motion and transfer 3D motion into motion tokens, similar to the generation process of word tokens. Building upon this "motion vocabulary", we perform language modeling on both motion and text in a unified manner, treating human motion as a specific language. Moreover, inspired by prompt learning, we pre-train MotionGPT with a mixture of motion-language data and fine-tune it on prompt-based question-and-answer tasks. Extensive experiments demonstrate that MotionGPT achieves state-of-the-art performances on multiple motion tasks including text-driven motion generation, motion captioning, motion prediction, and motion in-between.

Current approaches to learning semantic representations of sentences often use prior word-level knowledge. The current study aims to leverage visual information in order to capture sentence level semantics without the need for word embeddings. We use a multimodal sentence encoder trained on a corpus of images with matching text captions to produce visually grounded sentence embeddings. Deep Neural Networks are trained to map the two modalities to a common embedding space such that for an image the corresponding caption can be retrieved and vice versa. We show that our model achieves results comparable to the current state-of-the-art on two popular image-caption retrieval benchmark data sets: MSCOCO and Flickr8k. We evaluate the semantic content of the resulting sentence embeddings using the data from the Semantic Textual Similarity benchmark task and show that the multimodal embeddings correlate well with human semantic similarity judgements. The system achieves state-of-the-art results on several of these benchmarks, which shows that a system trained solely on multimodal data, without assuming any word representations, is able to capture sentence level semantics. Importantly, this result shows that we do not need prior knowledge of lexical level semantics in order to model sentence level semantics. These findings demonstrate the importance of visual information in semantics.

Recent video-text foundation models have demonstrated strong performance on a wide variety of downstream video understanding tasks. Can these video-text models genuinely understand the contents of natural videos? Standard video-text evaluations could be misleading as many questions can be inferred merely from the objects and contexts in a single frame or biases inherent in the datasets. In this paper, we aim to better assess the capabilities of current video-text models and understand their limitations. We propose a novel evaluation task for video-text understanding, namely retrieval from counterfactually augmented data (RCAD), and a new Feint6K dataset. To succeed on our new evaluation task, models must derive a comprehensive understanding of the video from cross-frame reasoning. Analyses show that previous video-text foundation models can be easily fooled by counterfactually augmented data and are far behind human-level performance. In order to narrow the gap between video-text models and human performance on RCAD, we identify a key limitation of current contrastive approaches on video-text data and introduce LLM-teacher, a more effective approach to learn action semantics by leveraging knowledge obtained from a pretrained large language model. Experiments and analyses show that our approach successfully learn more discriminative action embeddings and improves results on Feint6K when applied to multiple video-text models. Our Feint6K dataset and project page is available at https://feint6k.github.io.

Advancements in language foundation models have primarily fueled the recent surge in artificial intelligence. In contrast, generative learning of non-textual modalities, especially videos, significantly trails behind language modeling. This thesis chronicles our endeavor to build multi-task models for generating videos and other modalities under diverse conditions, as well as for understanding and compression applications. Given the high dimensionality of visual data, we pursue concise and accurate latent representations. Our video-native spatial-temporal tokenizers preserve high fidelity. We unveil a novel approach to mapping bidirectionally between visual observation and interpretable lexical terms. Furthermore, our scalable visual token representation proves beneficial across generation, compression, and understanding tasks. This achievement marks the first instances of language models surpassing diffusion models in visual synthesis and a video tokenizer outperforming industry-standard codecs. Within these multi-modal latent spaces, we study the design of multi-task generative models. Our masked multi-task transformer excels at the quality, efficiency, and flexibility of video generation. We enable a frozen language model, trained solely on text, to generate visual content. Finally, we build a scalable generative multi-modal transformer trained from scratch, enabling the generation of videos containing high-fidelity motion with the corresponding audio given diverse conditions. Throughout the course, we have shown the effectiveness of integrating multiple tasks, crafting high-fidelity latent representation, and generating multiple modalities. This work suggests intriguing potential for future exploration in generating non-textual data and enabling real-time, interactive experiences across various media forms.

Is vision good enough for language? Recent advancements in multimodal models primarily stem from the powerful reasoning abilities of large language models (LLMs). However, the visual component typically depends only on the instance-level contrastive language-image pre-training (CLIP). Our research reveals that the visual capabilities in recent multimodal LLMs (MLLMs) still exhibit systematic shortcomings. To understand the roots of these errors, we explore the gap between the visual embedding space of CLIP and vision-only self-supervised learning. We identify ''CLIP-blind pairs'' - images that CLIP perceives as similar despite their clear visual differences. With these pairs, we construct the Multimodal Visual Patterns (MMVP) benchmark. MMVP exposes areas where state-of-the-art systems, including GPT-4V, struggle with straightforward questions across nine basic visual patterns, often providing incorrect answers and hallucinated explanations. We further evaluate various CLIP-based vision-and-language models and found a notable correlation between visual patterns that challenge CLIP models and those problematic for multimodal LLMs. As an initial effort to address these issues, we propose a Mixture of Features (MoF) approach, demonstrating that integrating vision self-supervised learning features with MLLMs can significantly enhance their visual grounding capabilities. Together, our research suggests visual representation learning remains an open challenge, and accurate visual grounding is crucial for future successful multimodal systems.

Co-speech gestures play a vital role in non-verbal communication. In this paper, we introduce a new framework for co-speech gesture understanding in the wild. Specifically, we propose three new tasks and benchmarks to evaluate a model's capability to comprehend gesture-text-speech associations: (i) gesture-based retrieval, (ii) gestured word spotting, and (iii) active speaker detection using gestures. We present a new approach that learns a tri-modal speech-text-video-gesture representation to solve these tasks. By leveraging a combination of global phrase contrastive loss and local gesture-word coupling loss, we demonstrate that a strong gesture representation can be learned in a weakly supervised manner from videos in the wild. Our learned representations outperform previous methods, including large vision-language models (VLMs), across all three tasks. Further analysis reveals that speech and text modalities capture distinct gesture-related signals, underscoring the advantages of learning a shared tri-modal embedding space. The dataset, model, and code are available at: https://www.robots.ox.ac.uk/~vgg/research/jegal

Humans show an innate capability to identify tools to support specific actions. The association between objects parts and the actions they facilitate is usually named affordance. Being able to segment objects parts depending on the tasks they afford is crucial to enable intelligent robots to use objects of daily living. Traditional supervised learning methods for affordance segmentation require costly pixel-level annotations, while weakly supervised approaches, though less demanding, still rely on object-interaction examples and support a closed set of actions. These limitations hinder scalability, may introduce biases, and usually restrict models to a limited set of predefined actions. This paper proposes AffordanceCLIP, to overcome these limitations by leveraging the implicit affordance knowledge embedded within large pre-trained Vision-Language models like CLIP. We experimentally demonstrate that CLIP, although not explicitly trained for affordances detection, retains valuable information for the task. Our AffordanceCLIP achieves competitive zero-shot performance compared to methods with specialized training, while offering several advantages: i) it works with any action prompt, not just a predefined set; ii) it requires training only a small number of additional parameters compared to existing solutions and iii) eliminates the need for direct supervision on action-object pairs, opening new perspectives for functionality-based reasoning of models.

We present Magma, a foundation model that serves multimodal AI agentic tasks in both the digital and physical worlds. Magma is a significant extension of vision-language (VL) models in that it not only retains the VL understanding ability (verbal intelligence) of the latter, but is also equipped with the ability to plan and act in the visual-spatial world (spatial-temporal intelligence) and complete agentic tasks ranging from UI navigation to robot manipulation. To endow the agentic capabilities, Magma is pretrained on large amounts of heterogeneous datasets spanning from images, videos to robotics data, where the actionable visual objects (e.g., clickable buttons in GUI) in images are labeled by Set-of-Mark (SoM) for action grounding, and the object movements (e.g., the trace of human hands or robotic arms) in videos are labeled by Trace-of-Mark (ToM) for action planning. Extensive experiments show that SoM and ToM reach great synergy and facilitate the acquisition of spatial-temporal intelligence for our Magma model, which is fundamental to a wide range of tasks as shown in Fig.1. In particular, Magma creates new state-of-the-art results on UI navigation and robotic manipulation tasks, outperforming previous models that are specifically tailored to these tasks. On image and video-related multimodal tasks, Magma also compares favorably to popular large multimodal models that are trained on much larger datasets. We make our model and code public for reproducibility at https://microsoft.github.io/Magma.

Large multimodal models (LMMs) extend large language models (LLMs) with multi-sensory skills, such as visual understanding, to achieve stronger generic intelligence. In this paper, we analyze the latest model, GPT-4V(ision), to deepen the understanding of LMMs. The analysis focuses on the intriguing tasks that GPT-4V can perform, containing test samples to probe the quality and genericity of GPT-4V's capabilities, its supported inputs and working modes, and the effective ways to prompt the model. In our approach to exploring GPT-4V, we curate and organize a collection of carefully designed qualitative samples spanning a variety of domains and tasks. Observations from these samples demonstrate that GPT-4V's unprecedented ability in processing arbitrarily interleaved multimodal inputs and the genericity of its capabilities together make GPT-4V a powerful multimodal generalist system. Furthermore, GPT-4V's unique capability of understanding visual markers drawn on input images can give rise to new human-computer interaction methods such as visual referring prompting. We conclude the report with in-depth discussions on the emerging application scenarios and the future research directions for GPT-4V-based systems. We hope that this preliminary exploration will inspire future research on the next-generation multimodal task formulation, new ways to exploit and enhance LMMs to solve real-world problems, and gaining better understanding of multimodal foundation models.

MLLM agents demonstrate potential for complex embodied tasks by retrieving multimodal task-relevant trajectory data. However, current retrieval methods primarily focus on surface-level similarities of textual or visual cues in trajectories, neglecting their effectiveness for the specific task at hand. To address this issue, we propose a novel method, MLLM as ReTriever (MART), which enhances the performance of embodied agents by utilizing interaction data to fine-tune an MLLM retriever based on preference learning, such that the retriever fully considers the effectiveness of trajectories and prioritize them for unseen tasks. We also introduce Trajectory Abstraction, a mechanism that leverages MLLMs' summarization capabilities to represent trajectories with fewer tokens while preserving key information, enabling agents to better comprehend milestones in the trajectory. Experimental results across various environments demonstrate our method significantly improves task success rates in unseen scenes compared to baseline methods. This work presents a new paradigm for multimodal retrieval in embodied agents, by fine-tuning a general-purpose MLLM as the retriever to assess trajectory effectiveness. All benchmark task sets and simulator code modifications for action and observation spaces will be released.

Building joint representations across images and text is an essential step for tasks such as Visual Question Answering and Video Question Answering. In this work, we find that the representations must not only jointly capture features from both modalities but should also be diverse for better generalization performance. To this end, we propose joint vision-language representation learning by diversifying the tokenization learning process, enabling tokens that are sufficiently disentangled from each other to be learned from both modalities. We observe that our approach outperforms the baseline models in a majority of settings and is competitive with state-of-the-art methods.

We present M3P, a Multitask Multilingual Multimodal Pre-trained model that combines multilingual pre-training and multimodal pre-training into a unified framework via multitask pre-training. Our goal is to learn universal representations that can map objects occurred in different modalities or texts expressed in different languages into a common semantic space. In addition, to explicitly encourage fine-grained alignment between images and non-English languages, we also propose Multimodal Code-switched Training (MCT) to combine monolingual pre-training and multimodal pre-training via a code-switch strategy. Experiments are performed on the multilingual image retrieval task across two benchmark datasets, including MSCOCO and Multi30K. M3P can achieve comparable results for English and new state-of-the-art results for non-English languages.

We propose a novel framework for video understanding, called Temporally Contextualized CLIP (TC-CLIP), which leverages essential temporal information through global interactions in a spatio-temporal domain within a video. To be specific, we introduce Temporal Contextualization (TC), a layer-wise temporal information infusion mechanism for videos, which 1) extracts core information from each frame, 2) connects relevant information across frames for the summarization into context tokens, and 3) leverages the context tokens for feature encoding. Furthermore, the Video-conditional Prompting (VP) module processes context tokens to generate informative prompts in the text modality. Extensive experiments in zero-shot, few-shot, base-to-novel, and fully-supervised action recognition validate the effectiveness of our model. Ablation studies for TC and VP support our design choices. Our project page with the source code is available at https://github.com/naver-ai/tc-clip

We introduce a novel sequential modeling approach which enables learning a Large Vision Model (LVM) without making use of any linguistic data. To do this, we define a common format, "visual sentences", in which we can represent raw images and videos as well as annotated data sources such as semantic segmentations and depth reconstructions without needing any meta-knowledge beyond the pixels. Once this wide variety of visual data (comprising 420 billion tokens) is represented as sequences, the model can be trained to minimize a cross-entropy loss for next token prediction. By training across various scales of model architecture and data diversity, we provide empirical evidence that our models scale effectively. Many different vision tasks can be solved by designing suitable visual prompts at test time.

We present a new general learning approach for action recognition, Prompt Learning for Action Recognition (PLAR), which leverages the strengths of prompt learning to guide the learning process. Our approach is designed to predict the action label by helping the models focus on the descriptions or instructions associated with actions in the input videos. Our formulation uses various prompts, including optical flow, large vision models, and learnable prompts to improve the recognition performance. Moreover, we propose a learnable prompt method that learns to dynamically generate prompts from a pool of prompt experts under different inputs. By sharing the same objective, our proposed PLAR can optimize prompts that guide the model's predictions while explicitly learning input-invariant (prompt experts pool) and input-specific (data-dependent) prompt knowledge. We evaluate our approach on datasets consisting of both ground camera videos and aerial videos, and scenes with single-agent and multi-agent actions. In practice, we observe a 3.17-10.2% accuracy improvement on the aerial multi-agent dataset, Okutamam and 0.8-2.6% improvement on the ground camera single-agent dataset, Something Something V2. We plan to release our code on the WWW.

This paper reveals that large language models (LLMs), despite being trained solely on textual data, are surprisingly strong encoders for purely visual tasks in the absence of language. Even more intriguingly, this can be achieved by a simple yet previously overlooked strategy -- employing a frozen transformer block from pre-trained LLMs as a constituent encoder layer to directly process visual tokens. Our work pushes the boundaries of leveraging LLMs for computer vision tasks, significantly departing from conventional practices that typically necessitate a multi-modal vision-language setup with associated language prompts, inputs, or outputs. We demonstrate that our approach consistently enhances performance across a diverse range of tasks, encompassing pure 2D and 3D visual recognition tasks (e.g., image and point cloud classification), temporal modeling tasks (e.g., action recognition), non-semantic tasks (e.g., motion forecasting), and multi-modal tasks (e.g., 2D/3D visual question answering and image-text retrieval). Such improvements are a general phenomenon, applicable to various types of LLMs (e.g., LLaMA and OPT) and different LLM transformer blocks. We additionally propose the information filtering hypothesis to explain the effectiveness of pre-trained LLMs in visual encoding -- the pre-trained LLM transformer blocks discern informative visual tokens and further amplify their effect. This hypothesis is empirically supported by the observation that the feature activation, after training with LLM transformer blocks, exhibits a stronger focus on relevant regions. We hope that our work inspires new perspectives on utilizing LLMs and deepening our understanding of their underlying mechanisms. Code is available at https://github.com/ziqipang/LM4VisualEncoding.

Although large vision-language-action (VLA) models pretrained on extensive robot datasets offer promising generalist policies for robotic learning, they still struggle with spatial-temporal dynamics in interactive robotics, making them less effective in handling complex tasks, such as manipulation. In this work, we introduce visual trace prompting, a simple yet effective approach to facilitate VLA models' spatial-temporal awareness for action prediction by encoding state-action trajectories visually. We develop a new TraceVLA model by finetuning OpenVLA on our own collected dataset of 150K robot manipulation trajectories using visual trace prompting. Evaluations of TraceVLA across 137 configurations in SimplerEnv and 4 tasks on a physical WidowX robot demonstrate state-of-the-art performance, outperforming OpenVLA by 10% on SimplerEnv and 3.5x on real-robot tasks and exhibiting robust generalization across diverse embodiments and scenarios. To further validate the effectiveness and generality of our method, we present a compact VLA model based on 4B Phi-3-Vision, pretrained on the Open-X-Embodiment and finetuned on our dataset, rivals the 7B OpenVLA baseline while significantly improving inference efficiency.

Recent work in visual representation learning for robotics demonstrates the viability of learning from large video datasets of humans performing everyday tasks. Leveraging methods such as masked autoencoding and contrastive learning, these representations exhibit strong transfer to policy learning for visuomotor control. But, robot learning encompasses a diverse set of problems beyond control including grasp affordance prediction, language-conditioned imitation learning, and intent scoring for human-robot collaboration, amongst others. First, we demonstrate that existing representations yield inconsistent results across these tasks: masked autoencoding approaches pick up on low-level spatial features at the cost of high-level semantics, while contrastive learning approaches capture the opposite. We then introduce Voltron, a framework for language-driven representation learning from human videos and associated captions. Voltron trades off language-conditioned visual reconstruction to learn low-level visual patterns, and visually-grounded language generation to encode high-level semantics. We also construct a new evaluation suite spanning five distinct robot learning problems x2013 a unified platform for holistically evaluating visual representations for robotics. Through comprehensive, controlled experiments across all five problems, we find that Voltron's language-driven representations outperform the prior state-of-the-art, especially on targeted problems requiring higher-level features.

We present a target-aware video diffusion model that generates videos from an input image in which an actor interacts with a specified target while performing a desired action. The target is defined by a segmentation mask and the desired action is described via a text prompt. Unlike existing controllable image-to-video diffusion models that often rely on dense structural or motion cues to guide the actor's movements toward the target, our target-aware model requires only a simple mask to indicate the target, leveraging the generalization capabilities of pretrained models to produce plausible actions. This makes our method particularly effective for human-object interaction (HOI) scenarios, where providing precise action guidance is challenging, and further enables the use of video diffusion models for high-level action planning in applications such as robotics. We build our target-aware model by extending a baseline model to incorporate the target mask as an additional input. To enforce target awareness, we introduce a special token that encodes the target's spatial information within the text prompt. We then fine-tune the model with our curated dataset using a novel cross-attention loss that aligns the cross-attention maps associated with this token with the input target mask. To further improve performance, we selectively apply this loss to the most semantically relevant transformer blocks and attention regions. Experimental results show that our target-aware model outperforms existing solutions in generating videos where actors interact accurately with the specified targets. We further demonstrate its efficacy in two downstream applications: video content creation and zero-shot 3D HOI motion synthesis.

Large-scale pretrained image-text models have shown incredible zero-shot performance in a handful of tasks, including video ones such as action recognition and text-to-video retrieval. However, these models have not been adapted to video, mainly because they do not account for the time dimension but also because video frames are different from the typical images (e.g., containing motion blur, and less sharpness). In this paper, we present a fine-tuning strategy to refine these large-scale pretrained image-text models for zero-shot video understanding tasks. We show that by carefully adapting these models we obtain considerable improvements on two zero-shot Action Recognition tasks and three zero-shot Text-to-video Retrieval tasks. The code is available at https://github.com/bryant1410/fitclip

Recent Multimodal Large Language Models (MLLMs) have demonstrated significant progress in perceiving and reasoning over multimodal inquiries, ushering in a new research era for foundation models. However, vision-language misalignment in MLLMs has emerged as a critical challenge, where the textual responses generated by these models are not factually aligned with the given text-image inputs. Existing efforts to address vision-language misalignment have focused on developing specialized vision-language connectors or leveraging visual instruction tuning from diverse domains. In this paper, we tackle this issue from a fundamental yet unexplored perspective by revisiting the core architecture of MLLMs. Most MLLMs are typically built on decoder-only LLMs consisting of a causal attention mechanism, which limits the ability of earlier modalities (e.g., images) to incorporate information from later modalities (e.g., text). To address this problem, we propose AKI, a novel MLLM that unlocks causal attention into modality-mutual attention (MMA) to enable image tokens to attend to text tokens. This simple yet effective design allows AKI to achieve superior performance in 12 multimodal understanding benchmarks (+7.2% on average) without introducing additional parameters and increasing training time. Our MMA design is intended to be generic, allowing for application across various modalities, and scalable to accommodate diverse multimodal scenarios. The code is publicly available at https://github.com/sony/aki, and we will release our AKI-4B model to encourage further advancements in MLLMs across various directions.

Recent studies have presented compelling evidence that large language models (LLMs) can equip embodied agents with the self-driven capability to interact with the world, which marks an initial step toward versatile robotics. However, these efforts tend to overlook the visual richness of open worlds, rendering the entire interactive process akin to "a blindfolded text-based game." Consequently, LLM-based agents frequently encounter challenges in intuitively comprehending their surroundings and producing responses that are easy to understand. In this paper, we propose Steve-Eye, an end-to-end trained large multimodal model designed to address this limitation. Steve-Eye integrates the LLM with a visual encoder which enables it to process visual-text inputs and generate multimodal feedback. In addition, we use a semi-automatic strategy to collect an extensive dataset comprising 850K open-world instruction pairs, empowering our model to encompass three essential functions for an agent: multimodal perception, foundational knowledge base, and skill prediction and planning. Lastly, we develop three open-world evaluation benchmarks, then carry out extensive experiments from a wide range of perspectives to validate our model's capability to strategically act and plan. Codes and datasets will be released.

Learning visual knowledge from massive weakly-labeled web videos has attracted growing research interests thanks to the large corpus of easily accessible video data on the Internet. However, for video action recognition, the action of interest might only exist in arbitrary clips of untrimmed web videos, resulting in high label noises in the temporal space. To address this issue, we introduce a new method for pre-training video action recognition models using queried web videos. Instead of trying to filter out, we propose to convert the potential noises in these queried videos to useful supervision signals by defining the concept of Sub-Pseudo Label (SPL). Specifically, SPL spans out a new set of meaningful "middle ground" label space constructed by extrapolating the original weak labels during video querying and the prior knowledge distilled from a teacher model. Consequently, SPL provides enriched supervision for video models to learn better representations. SPL is fairly simple and orthogonal to popular teacher-student self-training frameworks without extra training cost. We validate the effectiveness of our method on four video action recognition datasets and a weakly-labeled image dataset to study the generalization ability. Experiments show that SPL outperforms several existing pre-training strategies using pseudo-labels and the learned representations lead to competitive results when fine-tuning on HMDB-51 and UCF-101 compared with recent pre-training methods.

There has been growing sentiment recently that modern large multimodal models (LMMs) have addressed most of the key challenges related to short video comprehension. As a result, both academia and industry are gradually shifting their attention towards the more complex challenges posed by understanding long-form videos. However, is this really the case? Our studies indicate that LMMs still lack many fundamental reasoning capabilities even when dealing with short videos. We introduce Vinoground, a temporal counterfactual LMM evaluation benchmark encompassing 1000 short and natural video-caption pairs. We demonstrate that existing LMMs severely struggle to distinguish temporal differences between different actions and object transformations. For example, the best model GPT-4o only obtains ~50% on our text and video scores, showing a large gap compared to the human baseline of ~90%. All open-source multimodal models and CLIP-based models perform much worse, producing mostly random chance performance. Through this work, we shed light onto the fact that temporal reasoning in short videos is a problem yet to be fully solved. The dataset and evaluation code are available at https://vinoground.github.io.

Recent literature shows that large-scale language modeling provides excellent reusable sentence representations with both recurrent and self-attentive architectures. However, there has been less clarity on the commonalities and differences in the representational properties induced by the two architectures. It also has been shown that visual information serves as one of the means for grounding sentence representations. In this paper, we present a meta-study assessing the representational quality of models where the training signal is obtained from different modalities, in particular, language modeling, image features prediction, and both textual and multimodal machine translation. We evaluate textual and visual features of sentence representations obtained using predominant approaches on image retrieval and semantic textual similarity. Our experiments reveal that on moderate-sized datasets, a sentence counterpart in a target language or visual modality provides much stronger training signal for sentence representation than language modeling. Importantly, we observe that while the Transformer models achieve superior machine translation quality, representations from the recurrent neural network based models perform significantly better over tasks focused on semantic relevance.

We introduce Kosmos-2, a Multimodal Large Language Model (MLLM), enabling new capabilities of perceiving object descriptions (e.g., bounding boxes) and grounding text to the visual world. Specifically, we represent refer expressions as links in Markdown, i.e., ``[text span](bounding boxes)'', where object descriptions are sequences of location tokens. Together with multimodal corpora, we construct large-scale data of grounded image-text pairs (called GrIT) to train the model. In addition to the existing capabilities of MLLMs (e.g., perceiving general modalities, following instructions, and performing in-context learning), Kosmos-2 integrates the grounding capability into downstream applications. We evaluate Kosmos-2 on a wide range of tasks, including (i) multimodal grounding, such as referring expression comprehension, and phrase grounding, (ii) multimodal referring, such as referring expression generation, (iii) perception-language tasks, and (iv) language understanding and generation. This work lays out the foundation for the development of Embodiment AI and sheds light on the big convergence of language, multimodal perception, action, and world modeling, which is a key step toward artificial general intelligence. Data, demo, and pretrained models are available at https://aka.ms/kosmos-2.

The correlation between the vision and text is essential for video moment retrieval (VMR), however, existing methods heavily rely on separate pre-training feature extractors for visual and textual understanding. Without sufficient temporal boundary annotations, it is non-trivial to learn universal video-text alignments. In this work, we explore multi-modal correlations derived from large-scale image-text data to facilitate generalisable VMR. To address the limitations of image-text pre-training models on capturing the video changes, we propose a generic method, referred to as Visual-Dynamic Injection (VDI), to empower the model's understanding of video moments. Whilst existing VMR methods are focusing on building temporal-aware video features, being aware of the text descriptions about the temporal changes is also critical but originally overlooked in pre-training by matching static images with sentences. Therefore, we extract visual context and spatial dynamic information from video frames and explicitly enforce their alignments with the phrases describing video changes (e.g. verb). By doing so, the potentially relevant visual and motion patterns in videos are encoded in the corresponding text embeddings (injected) so to enable more accurate video-text alignments. We conduct extensive experiments on two VMR benchmark datasets (Charades-STA and ActivityNet-Captions) and achieve state-of-the-art performances. Especially, VDI yields notable advantages when being tested on the out-of-distribution splits where the testing samples involve novel scenes and vocabulary.

Autonomous embodied agents live on an Internet of multimedia websites. Can they hop around multimodal websites to complete complex user tasks? Existing benchmarks fail to assess them in a realistic, evolving environment for their embodiment across websites. To answer this question, we present MMInA, a multihop and multimodal benchmark to evaluate the embodied agents for compositional Internet tasks, with several appealing properties: 1) Evolving real-world multimodal websites. Our benchmark uniquely operates on evolving real-world websites, ensuring a high degree of realism and applicability to natural user tasks. Our data includes 1,050 human-written tasks covering various domains such as shopping and travel, with each task requiring the agent to autonomously extract multimodal information from web pages as observations; 2) Multihop web browsing. Our dataset features naturally compositional tasks that require information from or actions on multiple websites to solve, to assess long-range reasoning capabilities on web tasks; 3) Holistic evaluation. We propose a novel protocol for evaluating an agent's progress in completing multihop tasks. We experiment with both standalone (multimodal) language models and heuristic-based web agents. Extensive experiments demonstrate that while long-chain multihop web tasks are easy for humans, they remain challenging for state-of-the-art web agents. We identify that agents are more likely to fail on the early hops when solving tasks of more hops, which results in lower task success rates. To address this issue, we propose a simple memory augmentation approach replaying past action trajectories to reflect. Our method significantly improved both the single-hop and multihop web browsing abilities of agents. See our code and data at https://mmina.cliangyu.com

The focal point of egocentric video understanding is modelling hand-object interactions. Standard models, e.g. CNNs or Vision Transformers, which receive RGB frames as input perform well. However, their performance improves further by employing additional input modalities that provide complementary cues, such as object detections, optical flow, audio, etc. The added complexity of the modality-specific modules, on the other hand, makes these models impractical for deployment. The goal of this work is to retain the performance of such a multimodal approach, while using only the RGB frames as input at inference time. We demonstrate that for egocentric action recognition on the Epic-Kitchens and the Something-Something datasets, students which are taught by multimodal teachers tend to be more accurate and better calibrated than architecturally equivalent models trained on ground truth labels in a unimodal or multimodal fashion. We further adopt a principled multimodal knowledge distillation framework, allowing us to deal with issues which occur when applying multimodal knowledge distillation in a naive manner. Lastly, we demonstrate the achieved reduction in computational complexity, and show that our approach maintains higher performance with the reduction of the number of input views. We release our code at https://github.com/gorjanradevski/multimodal-distillation.

Language models have recently advanced into the realm of reasoning, yet it is through multimodal reasoning that we can fully unlock the potential to achieve more comprehensive, human-like cognitive capabilities. This survey provides a systematic overview of the recent multimodal reasoning approaches, categorizing them into two levels: language-centric multimodal reasoning and collaborative multimodal reasoning. The former encompasses one-pass visual perception and active visual perception, where vision primarily serves a supporting role in language reasoning. The latter involves action generation and state update within reasoning process, enabling a more dynamic interaction between modalities. Furthermore, we analyze the technical evolution of these methods, discuss their inherent challenges, and introduce key benchmark tasks and evaluation metrics for assessing multimodal reasoning performance. Finally, we provide insights into future research directions from the following two perspectives: (i) from visual-language reasoning to omnimodal reasoning and (ii) from multimodal reasoning to multimodal agents. This survey aims to provide a structured overview that will inspire further advancements in multimodal reasoning research.

Learning text-video embeddings usually requires a dataset of video clips with manually provided captions. However, such datasets are expensive and time consuming to create and therefore difficult to obtain on a large scale. In this work, we propose instead to learn such embeddings from video data with readily available natural language annotations in the form of automatically transcribed narrations. The contributions of this work are three-fold. First, we introduce HowTo100M: a large-scale dataset of 136 million video clips sourced from 1.22M narrated instructional web videos depicting humans performing and describing over 23k different visual tasks. Our data collection procedure is fast, scalable and does not require any additional manual annotation. Second, we demonstrate that a text-video embedding trained on this data leads to state-of-the-art results for text-to-video retrieval and action localization on instructional video datasets such as YouCook2 or CrossTask. Finally, we show that this embedding transfers well to other domains: fine-tuning on generic Youtube videos (MSR-VTT dataset) and movies (LSMDC dataset) outperforms models trained on these datasets alone. Our dataset, code and models will be publicly available at: www.di.ens.fr/willow/research/howto100m/.

Temporal action detection (TAD) involves the localization and classification of action instances within untrimmed videos. While standard TAD follows fully supervised learning with closed-set setting on large training data, recent zero-shot TAD methods showcase the promising open-set setting by leveraging large-scale contrastive visual-language (ViL) pretrained models. However, existing zero-shot TAD methods have limitations on how to properly construct the strong relationship between two interdependent tasks of localization and classification and adapt ViL model to video understanding. In this work, we present ZEETAD, featuring two modules: dual-localization and zero-shot proposal classification. The former is a Transformer-based module that detects action events while selectively collecting crucial semantic embeddings for later recognition. The latter one, CLIP-based module, generates semantic embeddings from text and frame inputs for each temporal unit. Additionally, we enhance discriminative capability on unseen classes by minimally updating the frozen CLIP encoder with lightweight adapters. Extensive experiments on THUMOS14 and ActivityNet-1.3 datasets demonstrate our approach's superior performance in zero-shot TAD and effective knowledge transfer from ViL models to unseen action categories.

Massive multi-modality datasets play a significant role in facilitating the success of large video-language models. However, current video-language datasets primarily provide text descriptions for visual frames, considering audio to be weakly related information. They usually overlook exploring the potential of inherent audio-visual correlation, leading to monotonous annotation within each modality instead of comprehensive and precise descriptions. Such ignorance results in the difficulty of multiple cross-modality studies. To fulfill this gap, we present MMTrail, a large-scale multi-modality video-language dataset incorporating more than 20M trailer clips with visual captions, and 2M high-quality clips with multimodal captions. Trailers preview full-length video works and integrate context, visual frames, and background music. In particular, the trailer has two main advantages: (1) the topics are diverse, and the content characters are of various types, e.g., film, news, and gaming. (2) the corresponding background music is custom-designed, making it more coherent with the visual context. Upon these insights, we propose a systemic captioning framework, achieving various modality annotations with more than 27.1k hours of trailer videos. Here, to ensure the caption retains music perspective while preserving the authority of visual context, we leverage the advanced LLM to merge all annotations adaptively. In this fashion, our MMtrail dataset potentially paves the path for fine-grained large multimodal-language model training. In experiments, we provide evaluation metrics and benchmark results on our dataset, demonstrating the high quality of our annotation and its effectiveness for model training.

Generating instructional images of human daily actions from an egocentric viewpoint serves a key step towards efficient skill transfer. In this paper, we introduce a novel problem -- egocentric action frame generation. The goal is to synthesize the action frame conditioning on the user prompt question and an input egocentric image that captures user's environment. Notably, existing egocentric datasets lack the detailed annotations that describe the execution of actions. Additionally, the diffusion-based image manipulation models fail to control the state change of an action within the corresponding egocentric image pixel space. To this end, we finetune a visual large language model (VLLM) via visual instruction tuning for curating the enriched action descriptions to address our proposed problem. Moreover, we propose to Learn EGOcentric (LEGO) action frame generation using image and text embeddings from VLLM as additional conditioning. We validate our proposed model on two egocentric datasets -- Ego4D and Epic-Kitchens. Our experiments show prominent improvement over prior image manipulation models in both quantitative and qualitative evaluation. We also conduct detailed ablation studies and analysis to provide insights on our method.

Generative recommendation (GR) is an emerging paradigm where user actions are tokenized into discrete token patterns and autoregressively generated as predictions. However, existing GR models tokenize each action independently, assigning the same fixed tokens to identical actions across all sequences without considering contextual relationships. This lack of context-awareness can lead to suboptimal performance, as the same action may hold different meanings depending on its surrounding context. To address this issue, we propose ActionPiece to explicitly incorporate context when tokenizing action sequences. In ActionPiece, each action is represented as a set of item features, which serve as the initial tokens. Given the action sequence corpora, we construct the vocabulary by merging feature patterns as new tokens, based on their co-occurrence frequency both within individual sets and across adjacent sets. Considering the unordered nature of feature sets, we further introduce set permutation regularization, which produces multiple segmentations of action sequences with the same semantics. Experiments on public datasets demonstrate that ActionPiece consistently outperforms existing action tokenization methods, improving NDCG@10 by 6.00% to 12.82%.

As mobile devices are becoming ubiquitous, regularly interacting with a variety of user interfaces (UIs) is a common aspect of daily life for many people. To improve the accessibility of these devices and to enable their usage in a variety of settings, building models that can assist users and accomplish tasks through the UI is vitally important. However, there are several challenges to achieve this. First, UI components of similar appearance can have different functionalities, making understanding their function more important than just analyzing their appearance. Second, domain-specific features like Document Object Model (DOM) in web pages and View Hierarchy (VH) in mobile applications provide important signals about the semantics of UI elements, but these features are not in a natural language format. Third, owing to a large diversity in UIs and absence of standard DOM or VH representations, building a UI understanding model with high coverage requires large amounts of training data. Inspired by the success of pre-training based approaches in NLP for tackling a variety of problems in a data-efficient way, we introduce a new pre-trained UI representation model called ActionBert. Our methodology is designed to leverage visual, linguistic and domain-specific features in user interaction traces to pre-train generic feature representations of UIs and their components. Our key intuition is that user actions, e.g., a sequence of clicks on different UI components, reveals important information about their functionality. We evaluate the proposed model on a wide variety of downstream tasks, ranging from icon classification to UI component retrieval based on its natural language description. Experiments show that the proposed ActionBert model outperforms multi-modal baselines across all downstream tasks by up to 15.5%.

This paper introduces a novel method for self-supervised video representation learning via feature prediction. In contrast to the previous methods that focus on future feature prediction, we argue that a supervisory signal arising from unobserved past frames is complementary to one that originates from the future frames. The rationale behind our method is to encourage the network to explore the temporal structure of videos by distinguishing between future and past given present observations. We train our model in a contrastive learning framework, where joint encoding of future and past provides us with a comprehensive set of temporal hard negatives via swapping. We empirically show that utilizing both signals enriches the learned representations for the downstream task of action recognition. It outperforms independent prediction of future and past.

This tutorial note summarizes the presentation on ``Large Multimodal Models: Towards Building and Surpassing Multimodal GPT-4'', a part of CVPR 2023 tutorial on ``Recent Advances in Vision Foundation Models''. The tutorial consists of three parts. We first introduce the background on recent GPT-like large models for vision-and-language modeling to motivate the research in instruction-tuned large multimodal models (LMMs). As a pre-requisite, we describe the basics of instruction-tuning in large language models, which is further extended to the multimodal space. Lastly, we illustrate how to build the minimum prototype of multimodal GPT-4 like models with the open-source resource, and review the recently emerged topics.

Contrastive learning-based vision-language pre-training approaches, such as CLIP, have demonstrated great success in many vision-language tasks. These methods achieve cross-modal alignment by encoding a matched image-text pair with similar feature embeddings, which are generated by aggregating information from visual patches and language tokens. However, direct aligning cross-modal information using such representations is challenging, as visual patches and text tokens differ in semantic levels and granularities. To alleviate this issue, we propose a Finite Discrete Tokens (FDT) based multimodal representation. FDT is a set of learnable tokens representing certain visual-semantic concepts. Both images and texts are embedded using shared FDT by first grounding multimodal inputs to FDT space and then aggregating the activated FDT representations. The matched visual and semantic concepts are enforced to be represented by the same set of discrete tokens by a sparse activation constraint. As a result, the granularity gap between the two modalities is reduced. Through both quantitative and qualitative analyses, we demonstrate that using FDT representations in CLIP-style models improves cross-modal alignment and performance in visual recognition and vision-language downstream tasks. Furthermore, we show that our method can learn more comprehensive representations, and the learned FDT capture meaningful cross-modal correspondence, ranging from objects to actions and attributes.

Large scale Vision-Language (VL) models have shown tremendous success in aligning representations between visual and text modalities. This enables remarkable progress in zero-shot recognition, image generation & editing, and many other exciting tasks. However, VL models tend to over-represent objects while paying much less attention to verbs, and require additional tuning on video data for best zero-shot action recognition performance. While previous work relied on large-scale, fully-annotated data, in this work we propose an unsupervised approach. We adapt a VL model for zero-shot and few-shot action recognition using a collection of unlabeled videos and an unpaired action dictionary. Based on that, we leverage Large Language Models and VL models to build a text bag for each unlabeled video via matching, text expansion and captioning. We use those bags in a Multiple Instance Learning setup to adapt an image-text backbone to video data. Although finetuned on unlabeled video data, our resulting models demonstrate high transferability to numerous unseen zero-shot downstream tasks, improving the base VL model performance by up to 14\%, and even comparing favorably to fully-supervised baselines in both zero-shot and few-shot video recognition transfer. The code will be released later at https://github.com/wlin-at/MAXI.

Human motion generation is a cut-edge area of research in generative computer vision, with promising applications in video creation, game development, and robotic manipulation. The recent Mamba architecture shows promising results in efficiently modeling long and complex sequences, yet two significant challenges remain: Firstly, directly applying Mamba to extended motion generation is ineffective, as the limited capacity of the implicit memory leads to memory decay. Secondly, Mamba struggles with multimodal fusion compared to Transformers, and lack alignment with textual queries, often confusing directions (left or right) or omitting parts of longer text queries. To address these challenges, our paper presents three key contributions: Firstly, we introduce KMM, a novel architecture featuring Key frame Masking Modeling, designed to enhance Mamba's focus on key actions in motion segments. This approach addresses the memory decay problem and represents a pioneering method in customizing strategic frame-level masking in SSMs. Additionally, we designed a contrastive learning paradigm for addressing the multimodal fusion problem in Mamba and improving the motion-text alignment. Finally, we conducted extensive experiments on the go-to dataset, BABEL, achieving state-of-the-art performance with a reduction of more than 57% in FID and 70% parameters compared to previous state-of-the-art methods. See project website: https://steve-zeyu-zhang.github.io/KMM

Existing multimodal-based human action recognition approaches are computationally intensive, limiting their deployment in real-time applications. In this work, we present a novel and efficient pose-driven attention-guided multimodal network (EPAM-Net) for action recognition in videos. Specifically, we propose eXpand temporal Shift (X-ShiftNet) convolutional architectures for RGB and pose streams to capture spatio-temporal features from RGB videos and their skeleton sequences. The X-ShiftNet tackles the high computational cost of the 3D CNNs by integrating the Temporal Shift Module (TSM) into an efficient 2D CNN, enabling efficient spatiotemporal learning. Then skeleton features are utilized to guide the visual network stream, focusing on keyframes and their salient spatial regions using the proposed spatial-temporal attention block. Finally, the predictions of the two streams are fused for final classification. The experimental results show that our method, with a significant reduction in floating-point operations (FLOPs), outperforms and competes with the state-of-the-art methods on NTU RGB-D 60, NTU RGB-D 120, PKU-MMD, and Toyota SmartHome datasets. The proposed EPAM-Net provides up to a 72.8x reduction in FLOPs and up to a 48.6x reduction in the number of network parameters. The code will be available at https://github.com/ahmed-nady/Multimodal-Action-Recognition.

The integration of Large Language Models (LLMs) with visual encoders has recently shown promising performance in visual understanding tasks, leveraging their inherent capability to comprehend and generate human-like text for visual reasoning. Given the diverse nature of visual data, MultiModal Large Language Models (MM-LLMs) exhibit variations in model designing and training for understanding images, short videos, and long videos. Our paper focuses on the substantial differences and unique challenges posed by long video understanding compared to static image and short video understanding. Unlike static images, short videos encompass sequential frames with both spatial and within-event temporal information, while long videos consist of multiple events with between-event and long-term temporal information. In this survey, we aim to trace and summarize the advancements of MM-LLMs from image understanding to long video understanding. We review the differences among various visual understanding tasks and highlight the challenges in long video understanding, including more fine-grained spatiotemporal details, dynamic events, and long-term dependencies. We then provide a detailed summary of the advancements in MM-LLMs in terms of model design and training methodologies for understanding long videos. Finally, we compare the performance of existing MM-LLMs on video understanding benchmarks of various lengths and discuss potential future directions for MM-LLMs in long video understanding.

The advancement of large Vision-Language-Action (VLA) models has significantly improved robotic manipulation in terms of language-guided task execution and generalization to unseen scenarios. While existing VLAs adapted from pretrained large Vision-Language-Models (VLM) have demonstrated promising generalizability, their task performance is still unsatisfactory as indicated by the low tasks success rates in different environments. In this paper, we present a new advanced VLA architecture derived from VLM. Unlike previous works that directly repurpose VLM for action prediction by simple action quantization, we propose a omponentized VLA architecture that has a specialized action module conditioned on VLM output. We systematically study the design of the action module and demonstrates the strong performance enhancement with diffusion action transformers for action sequence modeling, as well as their favorable scaling behaviors. We also conduct comprehensive experiments and ablation studies to evaluate the efficacy of our models with varied designs. The evaluation on 5 robot embodiments in simulation and real work shows that our model not only significantly surpasses existing VLAs in task performance and but also exhibits remarkable adaptation to new robots and generalization to unseen objects and backgrounds. It exceeds the average success rates of OpenVLA which has similar model size (7B) with ours by over 35% in simulated evaluation and 55% in real robot experiments. It also outperforms the large RT-2-X model (55B) by 18% absolute success rates in simulation. Code and models can be found on our project page (https://cogact.github.io/).

Existing foundation models typically process visual input as pixels and textual input as tokens, a paradigm that contrasts with human perception, where both modalities are processed in a unified manner. With the rise of embodied and agentic AI, where inputs primarily come from camera pixels, the need for a unified perception framework becomes increasingly evident. In this paper, we propose to unify all modalities (text, tables, code, diagrams, images, etc) as pixel inputs, i.e. "Perceive Everything as Pixels" (PEAP). We introduce PixelWorld, a novel evaluation suite that unifies all the mentioned modalities into pixel space to gauge the existing models' performance. Our findings show that (1) PEAP outperforms baseline with token-based input in multimodal datasets, benefiting from unified input for better disambiguation, (2) significant declines in reasoning and coding capabilities across all models when processing pixel-based input, underscoring the need to enhance foundation models' perceptual abilities, (3) larger models can maintain strong performance on non-reasoning tasks under PEAP, while smaller models like Phi-3.5-V suffer significant performance degradation, (4) the attention pattern of PEAP is highly aligned with text token input, (5) PEAP can be accelerated significantly by exploiting the spatial sparsity. We conclude that the existing frontier models are competent in pixel perception, however, there is still headroom for improvement. Our code, dataset will be released upon acceptance.

In this paper, we study a novel problem in egocentric action recognition, which we term as "Multimodal Generalization" (MMG). MMG aims to study how systems can generalize when data from certain modalities is limited or even completely missing. We thoroughly investigate MMG in the context of standard supervised action recognition and the more challenging few-shot setting for learning new action categories. MMG consists of two novel scenarios, designed to support security, and efficiency considerations in real-world applications: (1) missing modality generalization where some modalities that were present during the train time are missing during the inference time, and (2) cross-modal zero-shot generalization, where the modalities present during the inference time and the training time are disjoint. To enable this investigation, we construct a new dataset MMG-Ego4D containing data points with video, audio, and inertial motion sensor (IMU) modalities. Our dataset is derived from Ego4D dataset, but processed and thoroughly re-annotated by human experts to facilitate research in the MMG problem. We evaluate a diverse array of models on MMG-Ego4D and propose new methods with improved generalization ability. In particular, we introduce a new fusion module with modality dropout training, contrastive-based alignment training, and a novel cross-modal prototypical loss for better few-shot performance. We hope this study will serve as a benchmark and guide future research in multimodal generalization problems. The benchmark and code will be available at https://github.com/facebookresearch/MMG_Ego4D.

We use multilayer Long Short Term Memory (LSTM) networks to learn representations of video sequences. Our model uses an encoder LSTM to map an input sequence into a fixed length representation. This representation is decoded using single or multiple decoder LSTMs to perform different tasks, such as reconstructing the input sequence, or predicting the future sequence. We experiment with two kinds of input sequences - patches of image pixels and high-level representations ("percepts") of video frames extracted using a pretrained convolutional net. We explore different design choices such as whether the decoder LSTMs should condition on the generated output. We analyze the outputs of the model qualitatively to see how well the model can extrapolate the learned video representation into the future and into the past. We try to visualize and interpret the learned features. We stress test the model by running it on longer time scales and on out-of-domain data. We further evaluate the representations by finetuning them for a supervised learning problem - human action recognition on the UCF-101 and HMDB-51 datasets. We show that the representations help improve classification accuracy, especially when there are only a few training examples. Even models pretrained on unrelated datasets (300 hours of YouTube videos) can help action recognition performance.

To fully leverage the capabilities of mobile manipulation robots, it is imperative that they are able to autonomously execute long-horizon tasks in large unexplored environments. While large language models (LLMs) have shown emergent reasoning skills on arbitrary tasks, existing work primarily concentrates on explored environments, typically focusing on either navigation or manipulation tasks in isolation. In this work, we propose MoMa-LLM, a novel approach that grounds language models within structured representations derived from open-vocabulary scene graphs, dynamically updated as the environment is explored. We tightly interleave these representations with an object-centric action space. The resulting approach is zero-shot, open-vocabulary, and readily extendable to a spectrum of mobile manipulation and household robotic tasks. We demonstrate the effectiveness of MoMa-LLM in a novel semantic interactive search task in large realistic indoor environments. In extensive experiments in both simulation and the real world, we show substantially improved search efficiency compared to conventional baselines and state-of-the-art approaches, as well as its applicability to more abstract tasks. We make the code publicly available at http://moma-llm.cs.uni-freiburg.de.

In this paper, we re-examine the task of cross-modal clip-sentence retrieval, where the clip is part of a longer untrimmed video. When the clip is short or visually ambiguous, knowledge of its local temporal context (i.e. surrounding video segments) can be used to improve the retrieval performance. We propose Context Transformer (ConTra); an encoder architecture that models the interaction between a video clip and its local temporal context in order to enhance its embedded representations. Importantly, we supervise the context transformer using contrastive losses in the cross-modal embedding space. We explore context transformers for video and text modalities. Results consistently demonstrate improved performance on three datasets: YouCook2, EPIC-KITCHENS and a clip-sentence version of ActivityNet Captions. Exhaustive ablation studies and context analysis show the efficacy of the proposed method.

Information retrieval is an ever-evolving and crucial research domain. The substantial demand for high-quality human motion data especially in online acquirement has led to a surge in human motion research works. Prior works have mainly concentrated on dual-modality learning, such as text and motion tasks, but three-modality learning has been rarely explored. Intuitively, an extra introduced modality can enrich a model's application scenario, and more importantly, an adequate choice of the extra modality can also act as an intermediary and enhance the alignment between the other two disparate modalities. In this work, we introduce LAVIMO (LAnguage-VIdeo-MOtion alignment), a novel framework for three-modality learning integrating human-centric videos as an additional modality, thereby effectively bridging the gap between text and motion. Moreover, our approach leverages a specially designed attention mechanism to foster enhanced alignment and synergistic effects among text, video, and motion modalities. Empirically, our results on the HumanML3D and KIT-ML datasets show that LAVIMO achieves state-of-the-art performance in various motion-related cross-modal retrieval tasks, including text-to-motion, motion-to-text, video-to-motion and motion-to-video.

This paper presents VDAct, a dataset for a Video-grounded Dialogue on Event-driven Activities, alongside VDEval, a session-based context evaluation metric specially designed for the task. Unlike existing datasets, VDAct includes longer and more complex video sequences that depict a variety of event-driven activities that require advanced contextual understanding for accurate response generation. The dataset comprises 3,000 dialogues with over 30,000 question-and-answer pairs, derived from 1,000 videos with diverse activity scenarios. VDAct displays a notably challenging characteristic due to its broad spectrum of activity scenarios and wide range of question types. Empirical studies on state-of-the-art vision foundation models highlight their limitations in addressing certain question types on our dataset. Furthermore, VDEval, which integrates dialogue session history and video content summaries extracted from our supplementary Knowledge Graphs to evaluate individual responses, demonstrates a significantly higher correlation with human assessments on the VDAct dataset than existing evaluation metrics that rely solely on the context of single dialogue turns.

This book is the result of a seminar in which we reviewed multimodal approaches and attempted to create a solid overview of the field, starting with the current state-of-the-art approaches in the two subfields of Deep Learning individually. Further, modeling frameworks are discussed where one modality is transformed into the other, as well as models in which one modality is utilized to enhance representation learning for the other. To conclude the second part, architectures with a focus on handling both modalities simultaneously are introduced. Finally, we also cover other modalities as well as general-purpose multi-modal models, which are able to handle different tasks on different modalities within one unified architecture. One interesting application (Generative Art) eventually caps off this booklet.

With the surge in the development of large language models, embodied intelligence has attracted increasing attention. Nevertheless, prior works on embodied intelligence typically encode scene or historical memory in an unimodal manner, either visual or linguistic, which complicates the alignment of the model's action planning with embodied control. To overcome this limitation, we introduce the Multimodal Embodied Interactive Agent (MEIA), capable of translating high-level tasks expressed in natural language into a sequence of executable actions. Specifically, we propose a novel Multimodal Environment Memory (MEM) module, facilitating the integration of embodied control with large models through the visual-language memory of scenes. This capability enables MEIA to generate executable action plans based on diverse requirements and the robot's capabilities. We conduct experiments in a dynamic virtual cafe environment, utilizing multiple large models through zero-shot learning, and carefully design scenarios for various situations. The experimental results showcase the promising performance of our MEIA in various embodied interactive tasks.

The human ability to easily solve multimodal tasks in context (i.e., with only a few demonstrations or simple instructions), is what current multimodal systems have largely struggled to imitate. In this work, we demonstrate that the task-agnostic in-context learning capabilities of large multimodal models can be significantly enhanced by effective scaling-up. We introduce Emu2, a generative multimodal model with 37 billion parameters, trained on large-scale multimodal sequences with a unified autoregressive objective. Emu2 exhibits strong multimodal in-context learning abilities, even emerging to solve tasks that require on-the-fly reasoning, such as visual prompting and object-grounded generation. The model sets a new record on multiple multimodal understanding tasks in few-shot settings. When instruction-tuned to follow specific instructions, Emu2 further achieves new state-of-the-art on challenging tasks such as question answering benchmarks for large multimodal models and open-ended subject-driven generation. These achievements demonstrate that Emu2 can serve as a base model and general-purpose interface for a wide range of multimodal tasks. Code and models are publicly available to facilitate future research.

In this work, we present an approach to construct a video-based robot policy capable of reliably executing diverse tasks across different robots and environments from few video demonstrations without using any action annotations. Our method leverages images as a task-agnostic representation, encoding both the state and action information, and text as a general representation for specifying robot goals. By synthesizing videos that ``hallucinate'' robot executing actions and in combination with dense correspondences between frames, our approach can infer the closed-formed action to execute to an environment without the need of any explicit action labels. This unique capability allows us to train the policy solely based on RGB videos and deploy learned policies to various robotic tasks. We demonstrate the efficacy of our approach in learning policies on table-top manipulation and navigation tasks. Additionally, we contribute an open-source framework for efficient video modeling, enabling the training of high-fidelity policy models with four GPUs within a single day.

We present a simplified, task-agnostic multi-modal pre-training approach that can accept either video or text input, or both for a variety of end tasks. Existing pre-training are task-specific by adopting either a single cross-modal encoder that requires both modalities, limiting their use for retrieval-style end tasks or more complex multitask learning with two unimodal encoders, limiting early cross-modal fusion. We instead introduce new pretraining masking schemes that better mix across modalities (e.g. by forcing masks for text to predict the closest video embeddings) while also maintaining separability (e.g. unimodal predictions are sometimes required, without using all the input). Experimental results show strong performance across a wider range of tasks than any previous methods, often outperforming task-specific pre-training. Code is made available at https://github.com/pytorch/fairseq/tree/main/examples/MMPT.

This study explores replacing Transformers in Visual Language Models (VLMs) with Mamba, a recent structured state space model (SSM) that demonstrates promising performance in sequence modeling. We test models up to 3B parameters under controlled conditions, showing that Mamba-based VLMs outperforms Transformers-based VLMs in captioning, question answering, and reading comprehension. However, we find that Transformers achieve greater performance in visual grounding and the performance gap widens with scale. We explore two hypotheses to explain this phenomenon: 1) the effect of task-agnostic visual encoding on the updates of the hidden states, and 2) the difficulty in performing visual grounding from the perspective of in-context multimodal retrieval. Our results indicate that a task-aware encoding yields minimal performance gains on grounding, however, Transformers significantly outperform Mamba at in-context multimodal retrieval. Overall, Mamba shows promising performance on tasks where the correct output relies on a summary of the image but struggles when retrieval of explicit information from the context is required.

Token-based world models emerged as a promising modular framework, modeling dynamics over token streams while optimizing tokenization separately. While successful in visual environments with discrete actions (e.g., Atari games), their broader applicability remains uncertain. In this paper, we introduce M^{3}, a modular world model that extends this framework, enabling flexible combinations of observation and action modalities through independent modality-specific components. M^{3} integrates several improvements from existing literature to enhance agent performance. Through extensive empirical evaluation across diverse benchmarks, M^{3} achieves state-of-the-art sample efficiency for planning-free world models. Notably, among these methods, it is the first to reach a human-level median score on Atari 100K, with superhuman performance on 13 games. We https://github.com/leor-c/M3{open-source our code and weights}.

Large Multimodal Models have demonstrated impressive capabilities in understanding general vision-language tasks. However, due to the limitation of supported input resolution (e.g., 448 x 448) as well as the inexhaustive description of the training image-text pair, these models often encounter challenges when dealing with intricate scene understandings and narratives. Here we address the problem by proposing the Monkey. Our contributions are two-fold: 1) without pretraining from the start, our method can be built upon an existing vision encoder (e.g., vit-BigHuge) to effectively improve the input resolution capacity up to 896 x 1344 pixels; 2) we propose a multi-level description generation method, which automatically provides rich information that can guide model to learn contextual association between scenes and objects. Our extensive testing across more than 16 distinct datasets reveals that Monkey achieves consistently competitive performance over the existing LMMs on fundamental tasks, such as Image Captioning, General Visual Question Answering (VQA), and Document-oriented VQA. Models, interactive demo, and the source code are provided at the following https://github.com/Yuliang-Liu/Monkey.

Video recordings of user activities, particularly desktop recordings, offer a rich source of data for understanding user behaviors and automating processes. However, despite advancements in Vision-Language Models (VLMs) and their increasing use in video analysis, extracting user actions from desktop recordings remains an underexplored area. This paper addresses this gap by proposing two novel VLM-based methods for user action extraction: the Direct Frame-Based Approach (DF), which inputs sampled frames directly into VLMs, and the Differential Frame-Based Approach (DiffF), which incorporates explicit frame differences detected via computer vision techniques. We evaluate these methods using a basic self-curated dataset and an advanced benchmark adapted from prior work. Our results show that the DF approach achieves an accuracy of 70% to 80% in identifying user actions, with the extracted action sequences being re-playable though Robotic Process Automation. We find that while VLMs show potential, incorporating explicit UI changes can degrade performance, making the DF approach more reliable. This work represents the first application of VLMs for extracting user action sequences from desktop recordings, contributing new methods, benchmarks, and insights for future research.

Controlling human motion based on text presents an important challenge in computer vision. Traditional approaches often rely on holistic action descriptions for motion synthesis, which struggle to capture subtle movements of local body parts. This limitation restricts the ability to isolate and manipulate specific movements. To address this, we propose a novel motion representation that decomposes motion into distinct body joint group movements and interactions from a kinematic perspective. We design an automatic dataset collection pipeline that enhances the existing text-motion benchmark by incorporating fine-grained local joint-group motion and interaction descriptions. To bridge the gap between text and motion domains, we introduce a hierarchical motion semantics approach that progressively fuses joint-level interaction information into the global action-level semantics for modality alignment. With this hierarchy, we introduce a coarse-to-fine motion synthesis procedure for various generation and editing downstream applications. Our quantitative and qualitative experiments demonstrate that the proposed formulation enhances text-motion retrieval by improving joint-spatial understanding, and enables more precise joint-motion generation and control. Project Page: {\smallhttps://andypinxinliu.github.io/KinMo/}

This paper aims to improve the performance of video multimodal large language models (MLLM) via long and rich context (LRC) modeling. As a result, we develop a new version of InternVideo2.5 with a focus on enhancing the original MLLMs' ability to perceive fine-grained details and capture long-form temporal structure in videos. Specifically, our approach incorporates dense vision task annotations into MLLMs using direct preference optimization and develops compact spatiotemporal representations through adaptive hierarchical token compression. Experimental results demonstrate this unique design of LRC greatly improves the results of video MLLM in mainstream video understanding benchmarks (short & long), enabling the MLLM to memorize significantly longer video inputs (at least 6x longer than the original), and master specialized vision capabilities like object tracking and segmentation. Our work highlights the importance of multimodal context richness (length and fineness) in empowering MLLM's innate abilites (focus and memory), providing new insights for future research on video MLLM. Code and models are available at https://github.com/OpenGVLab/InternVideo/tree/main/InternVideo2.5

Pre-trained representations are becoming crucial for many NLP and perception tasks. While representation learning in NLP has transitioned to training on raw text without human annotations, visual and vision-language representations still rely heavily on curated training datasets that are expensive or require expert knowledge. For vision applications, representations are mostly learned using datasets with explicit class labels such as ImageNet or OpenImages. For vision-language, popular datasets like Conceptual Captions, MSCOCO, or CLIP all involve a non-trivial data collection (and cleaning) process. This costly curation process limits the size of datasets and hence hinders the scaling of trained models. In this paper, we leverage a noisy dataset of over one billion image alt-text pairs, obtained without expensive filtering or post-processing steps in the Conceptual Captions dataset. A simple dual-encoder architecture learns to align visual and language representations of the image and text pairs using a contrastive loss. We show that the scale of our corpus can make up for its noise and leads to state-of-the-art representations even with such a simple learning scheme. Our visual representation achieves strong performance when transferred to classification tasks such as ImageNet and VTAB. The aligned visual and language representations enables zero-shot image classification and also set new state-of-the-art results on Flickr30K and MSCOCO image-text retrieval benchmarks, even when compared with more sophisticated cross-attention models. The representations also enable cross-modality search with complex text and text + image queries.

In this work we explore a new approach for robots to teach themselves about the world simply by observing it. In particular we investigate the effectiveness of learning task-agnostic representations for continuous control tasks. We extend Time-Contrastive Networks (TCN) that learn from visual observations by embedding multiple frames jointly in the embedding space as opposed to a single frame. We show that by doing so, we are now able to encode both position and velocity attributes significantly more accurately. We test the usefulness of this self-supervised approach in a reinforcement learning setting. We show that the representations learned by agents observing themselves take random actions, or other agents perform tasks successfully, can enable the learning of continuous control policies using algorithms like Proximal Policy Optimization (PPO) using only the learned embeddings as input. We also demonstrate significant improvements on the real-world Pouring dataset with a relative error reduction of 39.4% for motion attributes and 11.1% for static attributes compared to the single-frame baseline. Video results are available at https://sites.google.com/view/actionablerepresentations .

Multimodal LLMs have reached remarkable levels of proficiency in understanding multimodal inputs, driving extensive research to develop increasingly powerful models. However, much less attention has been paid to understanding and explaining the underlying mechanisms of these models. Most existing explainability research examines these models only in their final states, overlooking the dynamic representational shifts that occur during training. In this work, we systematically analyze the evolution of hidden state representations to reveal how fine-tuning alters the internal structure of a model to specialize in new multimodal tasks. Using a concept-based approach, we map hidden states to interpretable visual and textual concepts, enabling us to trace changes in encoded concepts across modalities as training progresses. We also demonstrate the use of shift vectors to capture these concepts changes. These shift vectors allow us to recover fine-tuned concepts by shifting those in the original model. Finally, we explore the practical impact of our findings on model steering, showing that we can adjust multimodal LLMs behaviors without any training, such as modifying answer types, captions style, or biasing the model toward specific responses. Our work sheds light on how multimodal representations evolve through fine-tuning and offers a new perspective for interpreting model adaptation in multimodal tasks. The code for this project is publicly available at https://github.com/mshukor/xl-vlms.

Embodied AI models often employ off the shelf vision backbones like CLIP to encode their visual observations. Although such general purpose representations encode rich syntactic and semantic information about the scene, much of this information is often irrelevant to the specific task at hand. This introduces noise within the learning process and distracts the agent's focus from task-relevant visual cues. Inspired by selective attention in humans-the process through which people filter their perception based on their experiences, knowledge, and the task at hand-we introduce a parameter-efficient approach to filter visual stimuli for embodied AI. Our approach induces a task-conditioned bottleneck using a small learnable codebook module. This codebook is trained jointly to optimize task reward and acts as a task-conditioned selective filter over the visual observation. Our experiments showcase state-of-the-art performance for object goal navigation and object displacement across 5 benchmarks, ProcTHOR, ArchitecTHOR, RoboTHOR, AI2-iTHOR, and ManipulaTHOR. The filtered representations produced by the codebook are also able generalize better and converge faster when adapted to other simulation environments such as Habitat. Our qualitative analyses show that agents explore their environments more effectively and their representations retain task-relevant information like target object recognition while ignoring superfluous information about other objects. Code and pretrained models are available at our project website: https://embodied-codebook.github.io.

Video action detection (VAD) aims to detect actors and classify their actions in a video. We figure that VAD suffers more from classification rather than localization of actors. Hence, we analyze how prevailing methods form features for classification and find that they prioritize actor regions, yet often overlooking the essential contextual information necessary for accurate classification. Accordingly, we propose to reduce the bias toward actor and encourage paying attention to the context that is relevant to each action class. By assigning a class-dedicated query to each action class, our model can dynamically determine where to focus for effective classification. The proposed model demonstrates superior performance on three challenging benchmarks with significantly fewer parameters and less computation.

Vision-Language-Action (VLA) models have shown remarkable potential in visuomotor control and instruction comprehension through end-to-end learning processes. However, current VLA models face significant challenges: they are slow during inference and require extensive pre-training on large amounts of robotic data, making real-world deployment difficult. In this paper, we introduce a new family of compact vision-language-action models, called TinyVLA, which offers two key advantages over existing VLA models: (1) faster inference speeds, and (2) improved data efficiency, eliminating the need for pre-training stage. Our framework incorporates two essential components to build TinyVLA: (1) initializing the policy backbone with robust, high-speed multimodal models, and (2) integrating a diffusion policy decoder during fine-tuning to enable precise robot actions. We conducted extensive evaluations of TinyVLA in both simulation and on real robots, demonstrating that our approach significantly outperforms the state-of-the-art VLA model, OpenVLA, in terms of speed and data efficiency, while delivering comparable or superior performance. Additionally, TinyVLA exhibits strong generalization capabilities across various dimensions, including language instructions, novel objects, unseen positions, changes in object appearance, background variations, and environmental shifts, often matching or exceeding the performance of OpenVLA. We believe that \methodname offers an interesting perspective on utilizing pre-trained multimodal models for policy learning. Our project is at https://tiny-vla.github.io.

Next generation virtual assistants are envisioned to handle multimodal inputs (e.g., vision, memories of previous interactions, in addition to the user's utterances), and perform multimodal actions (e.g., displaying a route in addition to generating the system's utterance). We introduce Situated Interactive MultiModal Conversations (SIMMC) as a new direction aimed at training agents that take multimodal actions grounded in a co-evolving multimodal input context in addition to the dialog history. We provide two SIMMC datasets totalling ~13K human-human dialogs (~169K utterances) using a multimodal Wizard-of-Oz (WoZ) setup, on two shopping domains: (a) furniture (grounded in a shared virtual environment) and, (b) fashion (grounded in an evolving set of images). We also provide logs of the items appearing in each scene, and contextual NLU and coreference annotations, using a novel and unified framework of SIMMC conversational acts for both user and assistant utterances. Finally, we present several tasks within SIMMC as objective evaluation protocols, such as Structural API Prediction and Response Generation. We benchmark a collection of existing models on these SIMMC tasks as strong baselines, and demonstrate rich multimodal conversational interactions. Our data, annotations, code, and models are publicly available.

Recent progress in using recurrent neural networks (RNNs) for image description has motivated the exploration of their application for video description. However, while images are static, working with videos requires modeling their dynamic temporal structure and then properly integrating that information into a natural language description. In this context, we propose an approach that successfully takes into account both the local and global temporal structure of videos to produce descriptions. First, our approach incorporates a spatial temporal 3-D convolutional neural network (3-D CNN) representation of the short temporal dynamics. The 3-D CNN representation is trained on video action recognition tasks, so as to produce a representation that is tuned to human motion and behavior. Second we propose a temporal attention mechanism that allows to go beyond local temporal modeling and learns to automatically select the most relevant temporal segments given the text-generating RNN. Our approach exceeds the current state-of-art for both BLEU and METEOR metrics on the Youtube2Text dataset. We also present results on a new, larger and more challenging dataset of paired video and natural language descriptions.

Recent advances in vision-language models have significantly propelled video understanding. Existing datasets and tasks, however, have notable limitations. Most datasets are confined to short videos with limited events and narrow narratives. For example, datasets with instructional and egocentric videos often document the activities of one person in a single scene. Although some movie datasets offer richer content, they are often limited to short-term tasks, lack publicly available videos and frequently encounter data leakage given the use of movie forums and other resources in LLM training. To address the above limitations, we propose the Short Film Dataset (SFD) with 1,078 publicly available amateur movies, a wide variety of genres and minimal data leakage issues. SFD offers long-term story-oriented video tasks in the form of multiple-choice and open-ended question answering. Our extensive experiments emphasize the need for long-term reasoning to solve SFD tasks. Notably, we find strong signals in movie transcripts leading to the on-par performance of people and LLMs. We also show significantly lower performance of current models compared to people when using vision data alone.

We propose an approach to self-supervised representation learning based on maximizing mutual information between features extracted from multiple views of a shared context. For example, one could produce multiple views of a local spatio-temporal context by observing it from different locations (e.g., camera positions within a scene), and via different modalities (e.g., tactile, auditory, or visual). Or, an ImageNet image could provide a context from which one produces multiple views by repeatedly applying data augmentation. Maximizing mutual information between features extracted from these views requires capturing information about high-level factors whose influence spans multiple views -- e.g., presence of certain objects or occurrence of certain events. Following our proposed approach, we develop a model which learns image representations that significantly outperform prior methods on the tasks we consider. Most notably, using self-supervised learning, our model learns representations which achieve 68.1% accuracy on ImageNet using standard linear evaluation. This beats prior results by over 12% and concurrent results by 7%. When we extend our model to use mixture-based representations, segmentation behaviour emerges as a natural side-effect. Our code is available online: https://github.com/Philip-Bachman/amdim-public.

Connecting text and visual modalities plays an essential role in generative intelligence. For this reason, inspired by the success of large language models, significant research efforts are being devoted to the development of Multimodal Large Language Models (MLLMs). These models can seamlessly integrate visual and textual modalities, both as input and output, while providing a dialogue-based interface and instruction-following capabilities. In this paper, we provide a comprehensive review of recent visual-based MLLMs, analyzing their architectural choices, multimodal alignment strategies, and training techniques. We also conduct a detailed analysis of these models across a wide range of tasks, including visual grounding, image generation and editing, visual understanding, and domain-specific applications. Additionally, we compile and describe training datasets and evaluation benchmarks, conducting comparisons among existing models in terms of performance and computational requirements. Overall, this survey offers a comprehensive overview of the current state of the art, laying the groundwork for future MLLMs.

We present VideoCLIP, a contrastive approach to pre-train a unified model for zero-shot video and text understanding, without using any labels on downstream tasks. VideoCLIP trains a transformer for video and text by contrasting temporally overlapping positive video-text pairs with hard negatives from nearest neighbor retrieval. Our experiments on a diverse series of downstream tasks, including sequence-level text-video retrieval, VideoQA, token-level action localization, and action segmentation reveal state-of-the-art performance, surpassing prior work, and in some cases even outperforming supervised approaches. Code is made available at https://github.com/pytorch/fairseq/tree/main/examples/MMPT.

Language models demonstrate remarkable capacity to generalize representations learned in one modality to downstream tasks in other modalities. Can we trace this ability to individual neurons? We study the case where a frozen text transformer is augmented with vision using a self-supervised visual encoder and a single linear projection learned on an image-to-text task. Outputs of the projection layer are not immediately decodable into language describing image content; instead, we find that translation between modalities occurs deeper within the transformer. We introduce a procedure for identifying "multimodal neurons" that convert visual representations into corresponding text, and decoding the concepts they inject into the model's residual stream. In a series of experiments, we show that multimodal neurons operate on specific visual concepts across inputs, and have a systematic causal effect on image captioning.

We introduce a new task called Referring Atomic Video Action Recognition (RAVAR), aimed at identifying atomic actions of a particular person based on a textual description and the video data of this person. This task differs from traditional action recognition and localization, where predictions are delivered for all present individuals. In contrast, we focus on recognizing the correct atomic action of a specific individual, guided by text. To explore this task, we present the RefAVA dataset, containing 36,630 instances with manually annotated textual descriptions of the individuals. To establish a strong initial benchmark, we implement and validate baselines from various domains, e.g., atomic action localization, video question answering, and text-video retrieval. Since these existing methods underperform on RAVAR, we introduce RefAtomNet -- a novel cross-stream attention-driven method specialized for the unique challenges of RAVAR: the need to interpret a textual referring expression for the targeted individual, utilize this reference to guide the spatial localization and harvest the prediction of the atomic actions for the referring person. The key ingredients are: (1) a multi-stream architecture that connects video, text, and a new location-semantic stream, and (2) cross-stream agent attention fusion and agent token fusion which amplify the most relevant information across these streams and consistently surpasses standard attention-based fusion on RAVAR. Extensive experiments demonstrate the effectiveness of RefAtomNet and its building blocks for recognizing the action of the described individual. The dataset and code will be made publicly available at https://github.com/KPeng9510/RAVAR.

With the bloom of Large Language Models (LLMs), Multimodal Large Language Models (MLLMs) that incorporate LLMs with pre-trained vision models have recently demonstrated impressive performance across diverse vision-language tasks. However, they fall short to comprehend context involving multiple images. A primary reason for this shortcoming is that the visual features for each images are encoded individually by frozen encoders before feeding into the LLM backbone, lacking awareness of other images and the multimodal instructions. We term this issue as prior-LLM modality isolation and propose a two phase paradigm, browse-and-concentrate, to enable in-depth multimodal context fusion prior to feeding the features into LLMs. This paradigm initially "browses" through the inputs for essential insights, and then revisits the inputs to "concentrate" on crucial details, guided by these insights, to achieve a more comprehensive understanding of the multimodal inputs. Additionally, we develop training strategies specifically to enhance the understanding of multi-image inputs. Our method markedly boosts the performance on 7 multi-image scenarios, contributing to increments on average accuracy by 2.13% and 7.60% against strong MLLMs baselines with 3B and 11B LLMs, respectively.

Predicting multiple real-world tasks in a single model often requires a particularly diverse feature space. Multimodal (MM) models aim to extract the synergistic predictive potential of multiple data types to create a shared feature space with aligned semantic meaning across inputs of drastically varying sizes (i.e. images, text, sound). Most current MM architectures fuse these representations in parallel, which not only limits their interpretability but also creates a dependency on modality availability. We present MultiModN, a multimodal, modular network that fuses latent representations in a sequence of any number, combination, or type of modality while providing granular real-time predictive feedback on any number or combination of predictive tasks. MultiModN's composable pipeline is interpretable-by-design, as well as innately multi-task and robust to the fundamental issue of biased missingness. We perform four experiments on several benchmark MM datasets across 10 real-world tasks (predicting medical diagnoses, academic performance, and weather), and show that MultiModN's sequential MM fusion does not compromise performance compared with a baseline of parallel fusion. By simulating the challenging bias of missing not-at-random (MNAR), this work shows that, contrary to MultiModN, parallel fusion baselines erroneously learn MNAR and suffer catastrophic failure when faced with different patterns of MNAR at inference. To the best of our knowledge, this is the first inherently MNAR-resistant approach to MM modeling. In conclusion, MultiModN provides granular insights, robustness, and flexibility without compromising performance.

Zero-Shot Temporal Action Localization (ZS-TAL) seeks to identify and locate actions in untrimmed videos unseen during training. Existing ZS-TAL methods involve fine-tuning a model on a large amount of annotated training data. While effective, training-based ZS-TAL approaches assume the availability of labeled data for supervised learning, which can be impractical in some applications. Furthermore, the training process naturally induces a domain bias into the learned model, which may adversely affect the model's generalization ability to arbitrary videos. These considerations prompt us to approach the ZS-TAL problem from a radically novel perspective, relaxing the requirement for training data. To this aim, we introduce a novel method that performs Test-Time adaptation for Temporal Action Localization (T3AL). In a nutshell, T3AL adapts a pre-trained Vision and Language Model (VLM). T3AL operates in three steps. First, a video-level pseudo-label of the action category is computed by aggregating information from the entire video. Then, action localization is performed adopting a novel procedure inspired by self-supervised learning. Finally, frame-level textual descriptions extracted with a state-of-the-art captioning model are employed for refining the action region proposals. We validate the effectiveness of T3AL by conducting experiments on the THUMOS14 and the ActivityNet-v1.3 datasets. Our results demonstrate that T3AL significantly outperforms zero-shot baselines based on state-of-the-art VLMs, confirming the benefit of a test-time adaptation approach.

Vision-Language Models (VLMs) have shown promising capabilities in handling various multimodal tasks, yet they struggle in long-context scenarios, particularly in tasks involving videos, high-resolution images, or lengthy image-text documents. In our work, we first conduct an empirical analysis of the long-context capabilities of VLMs using our augmented long-context multimodal datasets. Our findings reveal that directly applying the positional encoding mechanism used for textual tokens to visual tokens is suboptimal, and VLM performance degrades sharply when the position encoding exceeds the model's context window. To address this, we propose Variable Visual Position Encoding (V2PE), a novel positional encoding approach that employs variable and smaller increments for visual tokens, enabling more efficient management of long multimodal sequences. Our experiments demonstrate the effectiveness of V2PE to enhances VLMs' ability to effectively understand and reason over long multimodal contexts. We further integrate V2PE with our augmented long-context multimodal datasets to fine-tune the open-source VLM, InternVL2. The fine-tuned model achieves strong performance on both standard and long-context multimodal tasks. Notably, when the sequence length of the training dataset is increased to 256K tokens, the model is capable of processing multimodal sequences up to 1M tokens, highlighting its potential for real-world long-context applications.

Weakly-supervised Temporal Action Localization (W-TAL) aims to classify and localize all action instances in an untrimmed video under only video-level supervision. However, without frame-level annotations, it is challenging for W-TAL methods to identify false positive action proposals and generate action proposals with precise temporal boundaries. In this paper, we present a Two-Stream Consensus Network (TSCN) to simultaneously address these challenges. The proposed TSCN features an iterative refinement training method, where a frame-level pseudo ground truth is iteratively updated, and used to provide frame-level supervision for improved model training and false positive action proposal elimination. Furthermore, we propose a new attention normalization loss to encourage the predicted attention to act like a binary selection, and promote the precise localization of action instance boundaries. Experiments conducted on the THUMOS14 and ActivityNet datasets show that the proposed TSCN outperforms current state-of-the-art methods, and even achieves comparable results with some recent fully-supervised methods.

It is challenging for humans -- particularly those living with physical disabilities -- to control high-dimensional, dexterous robots. Prior work explores learning embedding functions that map a human's low-dimensional inputs (e.g., via a joystick) to complex, high-dimensional robot actions for assistive teleoperation; however, a central problem is that there are many more high-dimensional actions than available low-dimensional inputs. To extract the correct action and maximally assist their human controller, robots must reason over their context: for example, pressing a joystick down when interacting with a coffee cup indicates a different action than when interacting with knife. In this work, we develop assistive robots that condition their latent embeddings on visual inputs. We explore a spectrum of visual encoders and show that incorporating object detectors pretrained on small amounts of cheap, easy-to-collect structured data enables i) accurately and robustly recognizing the current context and ii) generalizing control embeddings to new objects and tasks. In user studies with a high-dimensional physical robot arm, participants leverage this approach to perform new tasks with unseen objects. Our results indicate that structured visual representations improve few-shot performance and are subjectively preferred by users.

We introduce AirLetters, a new video dataset consisting of real-world videos of human-generated, articulated motions. Specifically, our dataset requires a vision model to predict letters that humans draw in the air. Unlike existing video datasets, accurate classification predictions for AirLetters rely critically on discerning motion patterns and on integrating long-range information in the video over time. An extensive evaluation of state-of-the-art image and video understanding models on AirLetters shows that these methods perform poorly and fall far behind a human baseline. Our work shows that, despite recent progress in end-to-end video understanding, accurate representations of complex articulated motions -- a task that is trivial for humans -- remains an open problem for end-to-end learning.

We present ImageBind, an approach to learn a joint embedding across six different modalities - images, text, audio, depth, thermal, and IMU data. We show that all combinations of paired data are not necessary to train such a joint embedding, and only image-paired data is sufficient to bind the modalities together. ImageBind can leverage recent large scale vision-language models, and extends their zero-shot capabilities to new modalities just by using their natural pairing with images. It enables novel emergent applications 'out-of-the-box' including cross-modal retrieval, composing modalities with arithmetic, cross-modal detection and generation. The emergent capabilities improve with the strength of the image encoder and we set a new state-of-the-art on emergent zero-shot recognition tasks across modalities, outperforming specialist supervised models. Finally, we show strong few-shot recognition results outperforming prior work, and that ImageBind serves as a new way to evaluate vision models for visual and non-visual tasks.

The ability to integrate context, including perceptual and temporal cues, plays a pivotal role in grounding the meaning of a linguistic utterance. In order to measure to what extent current vision-and-language models master this ability, we devise a new multimodal challenge, Image Retrieval from Contextual Descriptions (ImageCoDe). In particular, models are tasked with retrieving the correct image from a set of 10 minimally contrastive candidates based on a contextual description. As such, each description contains only the details that help distinguish between images. Because of this, descriptions tend to be complex in terms of syntax and discourse and require drawing pragmatic inferences. Images are sourced from both static pictures and video frames. We benchmark several state-of-the-art models, including both cross-encoders such as ViLBERT and bi-encoders such as CLIP, on ImageCoDe. Our results reveal that these models dramatically lag behind human performance: the best variant achieves an accuracy of 20.9 on video frames and 59.4 on static pictures, compared with 90.8 in humans. Furthermore, we experiment with new model variants that are better equipped to incorporate visual and temporal context into their representations, which achieve modest gains. Our hope is that ImageCoDE will foster progress in grounded language understanding by encouraging models to focus on fine-grained visual differences.