# SadTalker **Repository Path**: dingjianfei/SadTalker ## Basic Information - **Project Name**: SadTalker - **Description**: ้•œๅƒ - **Primary Language**: Python - **License**: MIT - **Default Branch**: main - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 1 - **Forks**: 1 - **Created**: 2023-04-03 - **Last Updated**: 2023-10-04 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README
            [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/Winfredy/SadTalker/blob/main/quick_demo.ipynb)       [![Hugging Face Spaces](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue)](https://huggingface.co/spaces/vinthony/SadTalker)
Wenxuan Zhang *,1,2   Xiaodong Cun *,2  Xuan Wang 3  Yong Zhang 2  Xi 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](https://user-images.githubusercontent.com/4397546/222490039-b1f6156b-bf00-405b-9fda-0c9a9156f991.gif) TL;DR:       single portrait image ๐Ÿ™Žโ€โ™‚๏ธ      +       audio ๐ŸŽค       =       talking head video ๐ŸŽž.
## ๐Ÿ”ฅ Highlight - ๐Ÿ”ฅ Beta version of the `full image mode` is online! checkout [here](https://github.com/Winfredy/SadTalker#beta-full-bodyimage-generation) for more details. | still | still + enhancer | [input image @bagbag1815](https://twitter.com/bagbag1815/status/1642754319094108161) | |:--------------------: |:--------------------: | :----: | | | | - ๐Ÿ”ฅ Several new mode, eg, `still mode`, `reference mode`, `resize mode` are online for better and custom applications. - ๐Ÿ”ฅ Happy to see our method is used in various talking or singing avatar, checkout these wonderful demos at [bilibili](https://search.bilibili.com/all?keyword=sadtalker&from_source=webtop_search&spm_id_from=333.1007&search_source=3 ) and [twitter #sadtalker](https://twitter.com/search?q=%23sadtalker&src=typed_query). ## ๐Ÿ“‹ Changelog (Previous changelog can be founded [here](docs/changlelog.md)) - __[2023.03.30]__: Launch beta version of the full body mode. - __[2023.03.30]__: Launch new feature: through using reference videos, our algorithm can generate videos with more natural eye blinking and some eyebrow movement. - __[2023.03.29]__: `resize mode` is online by `python infererence.py --preprocess resize`! Where we can produce a larger crop of the image as discussed in https://github.com/Winfredy/SadTalker/issues/35. - __[2023.03.29]__: local gradio demo is online! `python app.py` to start the demo. New `requirments.txt` is used to avoid the bugs in `librosa`. - __[2023.03.28]__: Online demo is launched in [![Hugging Face Spaces](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue)](https://huggingface.co/spaces/vinthony/SadTalker), thanks AK! ## ๐ŸŽผ Pipeline ![main_of_sadtalker](https://user-images.githubusercontent.com/4397546/222490596-4c8a2115-49a7-42ad-a2c3-3bb3288a5f36.png) > Our method uses the coefficients of 3DMM as intermediate motion representation. To this end, we first generate realistic 3D motion coefficients (facial expression ฮฒ, head pose ฯ) from audio, then these coefficients are used to implicitly modulate the 3D-aware face render for final video generation. ## ๐Ÿšง TODO
Previous TODOs - [x] Generating 2D face from a single Image. - [x] Generating 3D face from Audio. - [x] Generating 4D free-view talking examples from audio and a single image. - [x] Gradio/Colab Demo. - [x] Full body/image Generation.
- [ ] training code of each componments. - [ ] Audio-driven Anime Avatar. - [ ] interpolate ChatGPT for a conversation demo ๐Ÿค” - [ ] integrade with stable-diffusion-web-ui. (stay tunning!) https://user-images.githubusercontent.com/4397546/222513483-89161f58-83d0-40e4-8e41-96c32b47bd4e.mp4 ## โš™๏ธ Installation #### Dependence Installation
CLICK ME For Mannual Installation ```bash 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 ```
CLICK For Docker Installation A dockerfile are also provided by [@thegenerativegeneration](https://github.com/thegenerativegeneration) in [docker hub](https://hub.docker.com/repository/docker/wawa9000/sadtalker), which can be used directly as: ```bash docker run --gpus "all" --rm -v $(pwd):/host_dir wawa9000/sadtalker \ --driven_audio /host_dir/deyu.wav \ --source_image /host_dir/image.jpg \ --expression_scale 1.0 \ --still \ --result_dir /host_dir ```
#### Download Trained Models
CLICK ME You can run the following script to put all the models in the right place. ```bash bash scripts/download_models.sh ``` OR download our pre-trained model from [google drive](https://drive.google.com/drive/folders/1Wd88VDoLhVzYsQ30_qDVluQr_Xm46yHT?usp=sharing) or our [github release page](https://github.com/Winfredy/SadTalker/releases/tag/v0.0.1), and then, put it in ./checkpoints. | Model | Description | :--- | :---------- |checkpoints/auido2exp_00300-model.pth | Pre-trained ExpNet in Sadtalker. |checkpoints/auido2pose_00140-model.pth | Pre-trained PoseVAE in Sadtalker. |checkpoints/mapping_00229-model.pth.tar | Pre-trained MappingNet in Sadtalker. |checkpoints/facevid2vid_00189-model.pth.tar | Pre-trained face-vid2vid model from [the reappearance of face-vid2vid](https://github.com/zhanglonghao1992/One-Shot_Free-View_Neural_Talking_Head_Synthesis). |checkpoints/epoch_20.pth | Pre-trained 3DMM extractor in [Deep3DFaceReconstruction](https://github.com/microsoft/Deep3DFaceReconstruction). |checkpoints/wav2lip.pth | Highly accurate lip-sync model in [Wav2lip](https://github.com/Rudrabha/Wav2Lip). |checkpoints/shape_predictor_68_face_landmarks.dat | Face landmark model used in [dilb](http://dlib.net/). |checkpoints/BFM | 3DMM library file. |checkpoints/hub | Face detection models used in [face alignment](https://github.com/1adrianb/face-alignment).
## ๐Ÿ”ฎ Quick Start #### Generating 2D face from a single Image from default config. ```bash python inference.py --driven_audio --source_image ``` The results will be saved in `results/$SOME_TIMESTAMP/*.mp4`. Or a local gradio demo can be run by: ```bash python app.py ``` #### Advanced Configuration
Click Me | Name | Configuration | default | Explaination | |:------------- |:------------- |:----- | :------------- | | Enhance Mode | `--enhancer` | None | Using `gfpgan` or `RestoreFormer` to enhance the generated face via face restoration network | Still Mode | ` --still` | False | Using the same pose parameters as the original image, fewer head motion. | Expressive Mode | `--expression_scale` | 1.0 | a larger value will make the expression motion stronger. | save path | `--result_dir` |`./results` | The file will be save in the newer location. | preprocess | `--preprocess` | `crop` | Run and produce the results in the croped input image. Other choices: `resize`, where the images will be resized to the specific resolution. | ref Mode (eye) | `--ref_eyeblink` | None | A video path, where we borrow the eyeblink from this reference video to provide more natural eyebrow movement. | ref Mode (pose) | `--ref_pose` | None | A video path, where we borrow the pose from the head reference video. | 3D Mode | `--face3dvis` | False | Need additional installation. More details to generate the 3d face can be founded [here](docs/face3d.md). | free-view Mode | `--input_yaw`,
`--input_pitch`,
`--input_roll` | None | Genearting novel view or free-view 4D talking head from a single image. More details can be founded [here](https://github.com/Winfredy/SadTalker#generating-4d-free-view-talking-examples-from-audio-and-a-single-image).
#### Examples | basic | w/ still mode | w/ exp_scale 1.3 | w/ gfpgan | |:-------------: |:-------------: |:-------------: |:-------------: | | | | | | > Kindly ensure to activate the audio as the default audio playing is incompatible with GitHub. | Input, w/ reference video , reference video | |:-------------: | | ![free_view](docs/using_ref_video.gif)| | If the reference video is shorter than the input audio, we will loop the reference video . #### Generating 3D face from Audio | Input | Animated 3d face | |:-------------: | :-------------: | | | | > Kindly ensure to activate the audio as the default audio playing is incompatible with GitHub. #### Generating 4D free-view talking examples from audio and a single image We use `input_yaw`, `input_pitch`, `input_roll` to control head pose. For example, `--input_yaw -20 30 10` means the input head yaw degree changes from -20 to 30 and then changes from 30 to 10. ```bash python inference.py --driven_audio \ --source_image \ --result_dir \ --input_yaw -20 30 10 ``` | Results, Free-view results, Novel view results | |:-------------: | | ![free_view](docs/free_view_result.gif)| #### [Beta Application] Full body/image Generation Now, you can use `--still` to generate a natural full body video. You can add `enhancer` or `full_img_enhancer` to improve the quality of the generated video. However, if you add other mode, such as `ref_eyeblinking`, `ref_pose`, the result will be bad. We are still trying to fix this problem. ```bash python inference.py --driven_audio \ --source_image \ --result_dir \ --still \ --enhancer gfpgan ``` ## ๐Ÿ›Ž Citation If you find our work useful in your research, please consider citing: ```bibtex @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} } ``` ## ๐Ÿ’— Acknowledgements Facerender code borrows heavily from [zhanglonghao's reproduction of face-vid2vid](https://github.com/zhanglonghao1992/One-Shot_Free-View_Neural_Talking_Head_Synthesis) and [PIRender](https://github.com/RenYurui/PIRender). We thank the authors for sharing their wonderful code. In training process, We also use the model from [Deep3DFaceReconstruction](https://github.com/microsoft/Deep3DFaceReconstruction) and [Wav2lip](https://github.com/Rudrabha/Wav2Lip). We thank for their wonderful work. ## ๐Ÿฅ‚ Related Works - [StyleHEAT: One-Shot High-Resolution Editable Talking Face Generation via Pre-trained StyleGAN (ECCV 2022)](https://github.com/FeiiYin/StyleHEAT) - [CodeTalker: Speech-Driven 3D Facial Animation with Discrete Motion Prior (CVPR 2023)](https://github.com/Doubiiu/CodeTalker) - [VideoReTalking: Audio-based Lip Synchronization for Talking Head Video Editing In the Wild (SIGGRAPH Asia 2022)](https://github.com/vinthony/video-retalking) - [DPE: Disentanglement of Pose and Expression for General Video Portrait Editing (CVPR 2023)](https://github.com/Carlyx/DPE) - [3D GAN Inversion with Facial Symmetry Prior (CVPR 2023)](https://github.com/FeiiYin/SPI/) - [T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations (CVPR 2023)](https://github.com/Mael-zys/T2M-GPT) ## ๐Ÿ“ข 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](ai.com), logo font๏ผš[Montserrat Alternates ](https://fonts.google.com/specimen/Montserrat+Alternates?preview.text=SadTalker&preview.text_type=custom&query=mont). All the copyright demo images are from communities users or the geneartion from stable diffusion. Free free to contact us if you feel uncomfortable.