Skip to content

wanggang1987/fast_sadtalker

Repository files navigation

🔥 INFO

基于sadtalker项目,做了大量优化和性能提升:

1.多显卡并行计算:在设备上具备多张显卡时,使用DP并行计算。推荐使用2张4090,达到最高性价比。(有详尽的多显卡多设备测试报告)

2.降低ehance帧数,整合多种补帧算法可选

3.调整参数,降低动作幅度和频率,更符合真人动作

4.增加silent模式合成

5.提供了完整的python api代码

6.api中包括ehance帧数和补帧倍数的参数

7.大量的代码优化,bug修复

8.大量的性能优化

9.重新训练的模型,解决了侧身侧脸照片合成后,头身错位的问题

    Open In Colab   Hugging Face Spaces   sd webui-colab   Replicate

Wenxuan Zhang *,1,2Xiaodong Cun *,2Xuan Wang 3Yong Zhang 2Xi Shen 2
Yu Guo1 Ying Shan 2   Fei Wang 1

1 Xi'an Jiaotong University   2 Tencent AI Lab   3 Ant Group  

CVPR 2023

sadtalker

TL;DR:       single portrait image 🙎‍♂️      +       audio 🎤       =       talking head video 🎞.


🔥 Highlight

sadtalker-webui.mp4
  • 🔥 full image mode is online! checkout here for more details.
still+enhancer in v0.0.1 still + enhancer in v0.0.2 input image @bagbag1815
still_e_n.mp4
full_body_2.bus_chinese_enhanced.mp4
  • 🔥 Several new mode, eg, still mode, reference mode, resize mode are online for better and custom applications.

  • 🔥 Happy to see more community demos at bilibili, Youtube and twitter #sadtalker.

📋 Changelog (Previous changelog can be founded here)

  • [2023.06.12]: add more new features in WEBUI extension, see the discussion here.

  • [2023.06.05]: release a new 512 beta face model. Fixed some bugs and improve the performance.

  • [2023.04.15]: Adding automatic1111 colab by @camenduru, thanks for this awesome colab: sd webui-colab.

  • [2023.04.12]: adding a more detailed sd-webui installation document, fixed reinstallation problem.

  • [2023.04.12]: Fixed the sd-webui safe issues becasue of the 3rd packages, optimize the output path in sd-webui-extension.

  • [2023.04.08]: ❗️❗️❗️ In v0.0.2, we add a logo watermark to the generated video to prevent abusing since it is very realistic.

  • [2023.04.08]: v0.0.2, full image animation, adding baidu driver for download checkpoints. Optimizing the logic about enhancer.

🚧 TODO: See the Discussion OpenTalker/SadTalker#280

If you have any problem, please view our FAQ before opening an issue.

⚙️ 1. Installation.

Tutorials from communities: 中文windows教程 | 日本語コース

Linux:

  1. Installing anaconda, python and git.

  2. Creating the env and install the requirements.

git clone https://github.com/Winfredy/SadTalker.git

cd SadTalker 

conda create -n sadtalker python=3.8

conda activate sadtalker

pip install torch==1.12.1+cu113 torchvision==0.13.1+cu113 torchaudio==0.12.1 --extra-index-url https://download.pytorch.org/whl/cu113

conda install ffmpeg

pip install -r requirements.txt

### tts is optional for gradio demo. 
### pip install TTS
  1. Install Python 3.10.6, checking "Add Python to PATH".
  2. Install git manually (OR scoop install git via scoop).
  3. Install ffmpeg, following this instruction (OR using scoop install ffmpeg via scoop).
  4. Download our SadTalker repository, for example by running git clone https://github.com/Winfredy/SadTalker.git.
  5. Download the checkpoint and gfpgan below↓.
  6. Run start.bat from Windows Explorer as normal, non-administrator, user, a gradio WebUI demo will be started.

Macbook:

More tips about installnation on Macbook and the Docker file can be founded here

📥 2. Download Trained Models.

You can run the following script to put all the models in the right place.

bash scripts/download_models.sh

Other alternatives:

we also provide an offline patch (gfpgan/), thus, no model will be downloaded when generating.

Google Driver: download our pre-trained model from this link (main checkpoints) and gfpgan (offline patch)

Github Release Page: download all the files from the lastest github release page, and then, put it in ./checkpoints.

百度云盘: we provided the downloaded model in checkpoints, 提取码: sadt. And gfpgan, 提取码: sadt.

Model Details

Model explains:

New version
Model Description
checkpoints/mapping_00229-model.pth.tar Pre-trained MappingNet in Sadtalker.
checkpoints/mapping_00109-model.pth.tar Pre-trained MappingNet in Sadtalker.
checkpoints/SadTalker_V0.0.2_256.safetensors packaged sadtalker checkpoints of old version, 256 face render).
checkpoints/SadTalker_V0.0.2_512.safetensors packaged sadtalker checkpoints of old version, 512 face render).
gfpgan/weights Face detection and enhanced models used in facexlib and gfpgan.
Old version
Model Description
checkpoints/auido2exp_00300-model.pth Pre-trained ExpNet in Sadtalker.
checkpoints/auido2pose_00140-model.pth Pre-trained PoseVAE in Sadtalker.
checkpoints/mapping_00229-model.pth.tar Pre-trained MappingNet in Sadtalker.
checkpoints/mapping_00109-model.pth.tar Pre-trained MappingNet in Sadtalker.
checkpoints/facevid2vid_00189-model.pth.tar Pre-trained face-vid2vid model from the reappearance of face-vid2vid.
checkpoints/epoch_20.pth Pre-trained 3DMM extractor in Deep3DFaceReconstruction.
checkpoints/wav2lip.pth Highly accurate lip-sync model in Wav2lip.
checkpoints/shape_predictor_68_face_landmarks.dat Face landmark model used in dilb.
checkpoints/BFM 3DMM library file.
checkpoints/hub Face detection models used in face alignment.
gfpgan/weights Face detection and enhanced models used in facexlib and gfpgan.

The final folder will be shown as:

image

🔮 3. Quick Start (Best Practice).

WebUI Demos:

Online: Huggingface | SDWebUI-Colab | Colab

Local Autiomatic1111 stable-diffusion webui extension: please refer to Autiomatic1111 stable-diffusion webui docs.

Local gradio demo(highly recommanded!): Similar to our hugging-face demo can be run by:

## you need manually install TTS(https://github.com/coqui-ai/TTS) via `pip install tts` in advanced.
python app.py

Local gradio demo(highly recommanded!):

  • windows: just double click webui.bat, the requirements will be installed automatically.
  • Linux/Mac OS: run bash webui.sh to start the webui.

Manually usages:

Animating a portrait image from default config:
python inference.py --driven_audio <audio.wav> \
                    --source_image <video.mp4 or picture.png> \
                    --enhancer gfpgan 

The results will be saved in results/$SOME_TIMESTAMP/*.mp4.

Full body/image Generation:

Using --still to generate a natural full body video. You can add enhancer to improve the quality of the generated video.

python inference.py --driven_audio <audio.wav> \
                    --source_image <video.mp4 or picture.png> \
                    --result_dir <a file to store results> \
                    --still \
                    --preprocess full \
                    --enhancer gfpgan 

More examples and configuration and tips can be founded in the >>> best practice documents <<<.

🛎 Citation

If you find our work useful in your research, please consider citing:

@article{zhang2022sadtalker,
  title={SadTalker: Learning Realistic 3D Motion Coefficients for Stylized Audio-Driven Single Image Talking Face Animation},
  author={Zhang, Wenxuan and Cun, Xiaodong and Wang, Xuan and Zhang, Yong and Shen, Xi and Guo, Yu and Shan, Ying and Wang, Fei},
  journal={arXiv preprint arXiv:2211.12194},
  year={2022}
}

💗 Acknowledgements

Facerender code borrows heavily from zhanglonghao's reproduction of face-vid2vid and PIRender. We thank the authors for sharing their wonderful code. In training process, We also use the model from Deep3DFaceReconstruction and Wav2lip. We thank for their wonderful work.

See also these wonderful 3rd libraries we use:

🥂 Extensions:

🥂 Related Works

📢 Disclaimer

This is not an official product of Tencent. This repository can only be used for personal/research/non-commercial purposes.

LOGO: color and font suggestion: ChatGPT, logo font:Montserrat Alternates .

All the copyright of the demo images and audio are from communities users or the geneartion from stable diffusion. Free free to contact us if you feel uncomfortable.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published