AnimeGamer: Infinite Anime Life Simulation with Next Game State Prediction

Junhao Cheng1,2, Yuying Ge1,โœ‰, Yixiao Ge1, Jing Liao2, Ying Shan1
1ARC Lab, Tencent PCG, 2City University of Hong Kong

Project Page arXiv Static Badge

Introduction

Experience the endless adventure of infinite anime life with AnimeGamer! ๐Ÿคฉ

AnimeGamer is a pioneering model designed for infinite anime life simulation. It leverages Multimodal Large Language Models (MLLMs) to generate dynamic animation shots that depict character movements and updates to character states. By incorporating historical visual context, AnimeGamer ensures contextual consistency and engaging gameplay. AnimeGamer uses novel action-aware multimodal representations and a video diffusion model to produce high-quality video clips, creating an immersive and ever-evolving gaming experience.

Citation

If you find the work helpful, please consider citing:

@misc{cheng2025animegamerinfiniteanimelife,
      title={AnimeGamer: Infinite Anime Life Simulation with Next Game State Prediction}, 
      author={Junhao Cheng and Yuying Ge and Yixiao Ge and Jing Liao and Ying Shan},
      year={2025},
      eprint={2504.01014},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2504.01014}, 
}
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support