README.md

---
license: mit
library_name: diffusers
pipeline_tag: image-to-video
extra_gated_prompt: You agree to not use the model to conduct experiments that cause
  harm to human subjects.
extra_gated_fields:
  Name: text
  Institution: text
  Email of your institution: text
  Country: country
  Specific date: date_picker
  I want to use this model for:
    type: select
    options:
    - Research
    - Education
    - label: Other
      value: other
  I agree to use this model for non-commercial use ONLY: checkbox
---

# 🌅 DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking Head Video Generation

[![arXiv](https://img.shields.io/badge/Arxiv-2410.13726-b31b1b.svg?logo=arXiv)](https://arxiv.org/abs/2410.13726)
[![Demo Page](https://img.shields.io/badge/Demo_Page-blue)](https://hanbo-cheng.github.io/DAWN/)
[![zhihu](https://img.shields.io/badge/知乎-0079FF.svg?logo=zhihu&logoColor=white)](https://zhuanlan.zhihu.com/p/2253009511)
[![DAWN](https://img.shields.io/badge/Github-181717?style=plastic&logo=github&logoColor=white)](https://github.com/Hanbo-Cheng/DAWN-pytorch)

`./hubert_ckpt`: you need to download [hubert-large-ls960-ft](https://huggingface.co/facebook/hubert-large-ls960-ft/tree/main)

`./pbnet`: checkpoint for the Pose-Blink generation network 

`DAWN_128.pth`: checkpoint for the A2V-FDM with 128 * 128 resolution

`DAWN_256.pth`: checkpoint for the A2V-FDM with 256 * 256 resolution

`LFG_128_1000ep.pth`: checkpoint for the LFG with 128 * 128 resolution

`LFG_256_400ep.pth`: checkpoint for the LFG with 256 * 256 resolution