MovieDreamer

<p align="center"> <img src="assets/logo.png" height=200> </p> <hr> <div align="center">

MovieDreamer: Hierarchical Generation for Coherent Long Visual Sequences

<p align="center"> <a href="https://aim-uofa.github.io/MovieDreamer/"><b>📖 Project Page</b></a> | <a href="https://arxiv.org/abs/2407.16655"><b>📄 Paper Link</b></a> | <a href="https://huggingface.co/papers/2407.16655"><b>🤗 Huggingface Paper Link</b></a> | <a href="https://www.youtube.com/watch?v=aubRVOGrKLU"><b>📹 Demo</b></a> </p> </div>

MovieDreamer aims to provide a solution for long content generation. Our method is capable of generating both long story and long video with multiple characters well-preserved!

<div align="center"> <img src="assets/demo.png" height=500> </div>

📣 News

• [7/20/2024] Paper released!

😲 Results

Please refer to our homepage for more thrilling results!

🛠️ Setup

• Comming Soon...

🚩 Plan

• [x] Story Results and Video Results.
• [x] Paper Released.
• [ ] Source Code and Pretrained Weights of Diffusion AutoEncoder.
• [ ] Source Code and Pretrained Weights of Autoregressive Model.
• [ ] Source code of Video Generation.
• [ ] Source code of Data Pre-processing.
• [ ] Source code of Training.

<!-- --- -->

🎫 License

For academic use, this project is licensed under the 2-clause BSD License. For commercial use, please contact Chunhua Shen.

🖊️ Citation

@misc{zhao2024moviedreamerhierarchicalgenerationcoherent,
      title={MovieDreamer: Hierarchical Generation for Coherent Long Visual Sequence}, 
      author={Canyu Zhao and Mingyu Liu and Wen Wang and Jianlong Yuan and Hao Chen and Bo Zhang and Chunhua Shen},
      year={2024},
      eprint={2407.16655},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2407.16655}, 
}