Update README.md

README.md

@@ -8,19 +8,76 @@ metrics:
 tags:
 - multimodal
 pipeline_tag: video-text-to-text
+model-index:
+- name: InternVL2.5_HiCo_R16
+  results:
+  - task:
+      type: multimodal
+    dataset:
+      name: MLVU
+      type: mlvu
+    metrics:
+    - type: accuracy
+      value: 71.5
+      name: accuracy
+      verified: true
+  - task:
+      type: multimodal
+    dataset:
+      name: MVBench
+      type: mvbench
+    metrics:
+    - type: accuracy
+      value: 74.0
+      name: accuracy
+      verified: true
+  - task:
+      type: multimodal
+    dataset:
+      name: Perception Test
+      type: percepTest
+    metrics:
+    - type: accuracy
+      value: 71.4
+      name: accuracy
+      verified: true
+  - task:
+      type: multimodal
+    dataset:
+      name: LongVideoBench
+      type: longvideobench
+    metrics:
+    - type: accuracy
+      value: 59.6
+      name: accuracy
+      verified: true
+  - task:
+      type: multimodal
+    dataset:
+      name: VideoMME (w/o sub)
+      type: videomme
+    metrics:
+    - type: accuracy
+      value: 64.9
+      name: accuracy
+      verified: true
+
 ---
 
-# 📕InternVL_2_5_HiCo_R16 ⚡
+# 📕InternVL2.5_HiCo_R16⚡
 <!-- [\[📰 Blog\]](https://internvideo.github.io/blog/2024-12-31-VideoChat-Flash) -->
 [\[📂 GitHub\]](https://github.com/OpenGVLab/InternVideo/tree/main/InternVideo2.5)  
 [\[📜 Tech Report\]](https://arxiv.org/abs/2501.12386) 
 <!-- [\[🗨️ Chat Demo\]](https://huggingface.co/spaces/OpenGVLab/VideoChat-Flash) -->
 
- 
+ InternVideo2.5 is a video multimodal large language model (MLLM, built upoon InternVL2.5) enhanced with **long and rich context (LRC) modeling**. It significantly improves upon existing MLLMs by enhancing their ability to perceive fine-grained details and capture long-form temporal structures. We achieve this through dense vision task annotations using direct preference optimization (TPO) and compact spatiotemporal representations via adaptive hierarchical token compression (HiCo). This model is a variant of InternVideo2.5's ablation experiment, built on HiCo technology only (R16 means 16 tokens per frame).
+
+
+
 ## 📈 Performance
 | Model |  MVBench | LongVideoBench |  VideoMME(w/o sub)| 
 | ---   |  ---     |   ---            | ---     | 
-|InternVL_2_5_HiCo_R16| - |  -   | - | 
+|InternVL2.5_HiCo_R16| 74.0 |  59.6   | 64.9| 
 
 ## 🚀 How to use the model
 
@@ -35,16 +92,130 @@ pip install flash-attn --no-build-isolation
 ```
 Then you could use our model:
 ```python
+import numpy as np
+import torch
+import torchvision.transforms as T
+from decord import VideoReader, cpu
+from PIL import Image
+from torchvision.transforms.functional import InterpolationMode
 from transformers import AutoModel, AutoTokenizer
 
+
 # model setting
 model_path = 'OpenGVLab/InternVL_2_5_HiCo_R16'
 
 tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)
 model = AutoModel.from_pretrained(model_path, trust_remote_code=True).half().cuda()
-image_processor = model.get_vision_tower().image_processor
+
+IMAGENET_MEAN = (0.485, 0.456, 0.406)
+IMAGENET_STD = (0.229, 0.224, 0.225)
+
+def build_transform(input_size):
+    MEAN, STD = IMAGENET_MEAN, IMAGENET_STD
+    transform = T.Compose([T.Lambda(lambda img: img.convert("RGB") if img.mode != "RGB" else img), T.Resize((input_size, input_size), interpolation=InterpolationMode.BICUBIC), T.ToTensor(), T.Normalize(mean=MEAN, std=STD)])
+    return transform
+
+
+def find_closest_aspect_ratio(aspect_ratio, target_ratios, width, height, image_size):
+    best_ratio_diff = float("inf")
+    best_ratio = (1, 1)
+    area = width * height
+    for ratio in target_ratios:
+        target_aspect_ratio = ratio[0] / ratio[1]
+        ratio_diff = abs(aspect_ratio - target_aspect_ratio)
+        if ratio_diff < best_ratio_diff:
+            best_ratio_diff = ratio_diff
+            best_ratio = ratio
+        elif ratio_diff == best_ratio_diff:
+            if area > 0.5 * image_size * image_size * ratio[0] * ratio[1]:
+                best_ratio = ratio
+    return best_ratio
+
+
+def dynamic_preprocess(image, min_num=1, max_num=6, image_size=448, use_thumbnail=False):
+    orig_width, orig_height = image.size
+    aspect_ratio = orig_width / orig_height
+
+    # calculate the existing image aspect ratio
+    target_ratios = set((i, j) for n in range(min_num, max_num + 1) for i in range(1, n + 1) for j in range(1, n + 1) if i * j <= max_num and i * j >= min_num)
+    target_ratios = sorted(target_ratios, key=lambda x: x[0] * x[1])
+
+    # find the closest aspect ratio to the target
+    target_aspect_ratio = find_closest_aspect_ratio(aspect_ratio, target_ratios, orig_width, orig_height, image_size)
+
+    # calculate the target width and height
+    target_width = image_size * target_aspect_ratio[0]
+    target_height = image_size * target_aspect_ratio[1]
+    blocks = target_aspect_ratio[0] * target_aspect_ratio[1]
+
+    # resize the image
+    resized_img = image.resize((target_width, target_height))
+    processed_images = []
+    for i in range(blocks):
+        box = ((i % (target_width // image_size)) * image_size, (i // (target_width // image_size)) * image_size, ((i % (target_width // image_size)) + 1) * image_size, ((i // (target_width // image_size)) + 1) * image_size)
+        # split the image
+        split_img = resized_img.crop(box)
+        processed_images.append(split_img)
+    assert len(processed_images) == blocks
+    if use_thumbnail and len(processed_images) != 1:
+        thumbnail_img = image.resize((image_size, image_size))
+        processed_images.append(thumbnail_img)
+    return processed_images
 
 
+def load_image(image, input_size=448, max_num=6):
+    transform = build_transform(input_size=input_size)
+    images = dynamic_preprocess(image, image_size=input_size, use_thumbnail=True, max_num=max_num)
+    pixel_values = [transform(image) for image in images]
+    pixel_values = torch.stack(pixel_values)
+    return pixel_values
+
+
+def get_index(bound, fps, max_frame, first_idx=0, num_segments=32):
+    if bound:
+        start, end = bound[0], bound[1]
+    else:
+        start, end = -100000, 100000
+    start_idx = max(first_idx, round(start * fps))
+    end_idx = min(round(end * fps), max_frame)
+    seg_size = float(end_idx - start_idx) / num_segments
+    frame_indices = np.array([int(start_idx + (seg_size / 2) + np.round(seg_size * idx)) for idx in range(num_segments)])
+    return frame_indices
+
+def get_num_frames_by_duration(duration):
+        local_num_frames = 4        
+        num_segments = int(duration // local_num_frames)
+        if num_segments == 0:
+            num_frames = local_num_frames
+        else:
+            num_frames = local_num_frames * num_segments
+        
+        num_frames = min(512, num_frames)
+        num_frames = max(128, num_frames)
+
+        return num_frames
+
+def load_video(video_path, bound=None, input_size=448, max_num=1, num_segments=32, get_frame_by_duration = False):
+    vr = VideoReader(video_path, ctx=cpu(0), num_threads=1)
+    max_frame = len(vr) - 1
+    fps = float(vr.get_avg_fps())
+
+    pixel_values_list, num_patches_list = [], []
+    transform = build_transform(input_size=input_size)
+    if get_frame_by_duration:
+        duration = max_frame / fps
+        num_segments = get_num_frames_by_duration(duration)
+    frame_indices = get_index(bound, fps, max_frame, first_idx=0, num_segments=num_segments)
+    for frame_index in frame_indices:
+        img = Image.fromarray(vr[frame_index].asnumpy()).convert("RGB")
+        img = dynamic_preprocess(img, image_size=input_size, use_thumbnail=True, max_num=max_num)
+        pixel_values = [transform(tile) for tile in img]
+        pixel_values = torch.stack(pixel_values)
+        num_patches_list.append(pixel_values.shape[0])
+        pixel_values_list.append(pixel_values)
+    pixel_values = torch.cat(pixel_values_list)
+    return pixel_values, num_patches_list
+
 # evaluation setting
 max_num_frames = 512
 generation_config = dict(
@@ -54,20 +225,26 @@ generation_config = dict(
     top_p=0.1,
     num_beams=1
 )
-
 video_path = "your_video.mp4"
+num_segments=128
 
-# single-turn conversation
-question1 = "Describe this video in detail."
-output1, chat_history = model.chat(video_path=video_path, tokenizer=tokenizer, user_prompt=question1, return_history=True, max_num_frames=max_num_frames, generation_config=generation_config)
-
-print(output1)
 
-# multi-turn conversation
-question2 = "How many people appear in the video?"
-output2, chat_history = model.chat(video_path=video_path, tokenizer=tokenizer, user_prompt=question2, chat_history=chat_history, return_history=True, max_num_frames=max_num_frames, generation_config=generation_config)
-
-print(output2)
+with torch.no_grad():
+  
+  pixel_values, num_patches_list = load_video(video_path, num_segments=num_segments, max_num=1, get_frame_by_duration=False)
+  pixel_values = pixel_values.to(torch.bfloat16).to(model.device)
+  video_prefix = "".join([f"Frame{i+1}: <image>\n" for i in range(len(num_patches_list))])
+  # single-turn conversation
+  question1 = "Describe this video in detail."
+  question = video_prefix + question1
+  output1, chat_history = model.chat(tokenizer, pixel_values, question, generation_config, num_patches_list=num_patches_list, history=None, return_history=True)
+  print(output1)
+  
+  # multi-turn conversation
+  question2 = "How many people appear in the video?"
+  output2, chat_history = model.chat(tokenizer, pixel_values, question, generation_config, num_patches_list=num_patches_list, history=chat_history, return_history=True)
+  
+  print(output2)
 ```
 
 ## ✏️ Citation
@@ -80,4 +257,13 @@ print(output2)
   journal={arXiv preprint arXiv:2501.12386},
   year={2025}
 }
+
+
+@article{li2024videochatflash,
+  title={VideoChat-Flash: Hierarchical Compression for Long-Context Video Modeling},
+  author={Li, Xinhao and Wang, Yi and Yu, Jiashuo and Zeng, Xiangyu and Zhu, Yuhan and Huang, Haian and Gao, Jianfei and Li, Kunchang and He, Yinan and Wang, Chenting and others},
+  journal={arXiv preprint arXiv:2501.00574},
+  year={2024}
+}
+
 ```