A Web UI using Gradio for Wav2Lip
I wanted to build the UI with gradio. However, gradio requires python 3.8 while wav2lip requires 3.6. I ended up creating 2 conda environments. One with 3.6 for wav2lip and one with 3.8 for gradio, then had the gradio call a cmd script with input parameters selected from the Web UI and the cmd script change to the wav2lip 3.6 environment and call inferency.py with the provided parameters.
Video of the coding session where this was created: https://youtu.be/pYpcvArk_4U
Quick video of the installation: https://youtu.be/n3G9MVR6OJY
Open a conda prompt to a directory where you would like to install the wav2lip web ui.
conda create -n wav2lip python=3.10
conda activate wav2lip
git clone https://github.com/natlamir/Wav2Lip-WebUI.git wav2lip
cd wav2lip
conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
pip install -r requirements.txt
Remember to get the model and weights from below. With those things done, you should be able to double click the run.cmd to launch the Web UI.
Make sure to download the pre-trained model and weights and place them in the appropriate directory:
Pre-trained model (rename to s3fd.pth and place in face_detection/detection/sfd
folder):
https://www.adrianbulat.com/downloads/python-fan/s3fd-619a316812.pth
Weights (place in checkpoints
folder):
- https://iiitaphyd-my.sharepoint.com/:u:/g/personal/radrabha_m_research_iiit_ac_in/Eb3LEzbfuKlJiR600lQWRxgBIY27JZg80f7V9jtMfbNDaQ?e=TBFBVW
- https://iiitaphyd-my.sharepoint.com/:u:/g/personal/radrabha_m_research_iiit_ac_in/EdjI7bZlgApMqsVoEUUXpLsBxqXbn5z8VTmoxp55YNDcIA?e=n9ljGW
For commercial requests, please contact us at [email protected] or [email protected]. We have an HD model ready that can be used commercially.
This code is part of the paper: A Lip Sync Expert Is All You Need for Speech to Lip Generation In the Wild published at ACM Multimedia 2020.
๐ Original Paper | ๐ฐ Project Page | ๐ Demo | โก Live Testing | ๐ Colab Notebook |
---|---|---|---|---|
Paper | Project Page | Demo Video | Interactive Demo | Colab Notebook /Updated Collab Notebook |
- Weights of the visual quality disc has been updated in readme!
- Lip-sync videos to any target speech with high accuracy ๐ฏ. Try our interactive demo.
- โจ Works for any identity, voice, and language. Also works for CGI faces and synthetic voices.
- Complete training code, inference code, and pretrained models are available ๐ฅ
- Or, quick-start with the Google Colab Notebook: Link. Checkpoints and samples are available in a Google Drive folder as well. There is also a tutorial video on this, courtesy of What Make Art. Also, thanks to Eyal Gruss, there is a more accessible Google Colab notebook with more useful features. A tutorial collab notebook is present at this link.
- ๐ฅ ๐ฅ Several new, reliable evaluation benchmarks and metrics [
evaluation/
folder of this repo] released. Instructions to calculate the metrics reported in the paper are also present.
All results from this open-source code or our demo website should only be used for research/academic/personal purposes only. As the models are trained on the LRS2 dataset, any form of commercial use is strictly prohibited. For commercial requests please contact us directly!
Python 3.6
- ffmpeg:
sudo apt-get install ffmpeg
- Install necessary packages using
pip install -r requirements.txt
. Alternatively, instructions for using a docker image is provided here. Have a look at this comment and comment on the gist if you encounter any issues. - Face detection pre-trained model should be downloaded to
face_detection/detection/sfd/s3fd.pth
. Alternative link if the above does not work.
Model | Description | Link to the model |
---|---|---|
Wav2Lip | Highly accurate lip-sync | Link |
Wav2Lip + GAN | Slightly inferior lip-sync, but better visual quality | Link |
Expert Discriminator | Weights of the expert discriminator | Link |
Visual Quality Discriminator | Weights of the visual disc trained in a GAN setup | Link |
You can lip-sync any video to any audio:
python inference.py --checkpoint_path <ckpt> --face <video.mp4> --audio <an-audio-source>
The result is saved (by default) in results/result_voice.mp4
. You can specify it as an argument, similar to several other available options. The audio source can be any file supported by FFMPEG
containing audio data: *.wav
, *.mp3
or even a video file, from which the code will automatically extract the audio.
- Experiment with the
--pads
argument to adjust the detected face bounding box. Often leads to improved results. You might need to increase the bottom padding to include the chin region. E.g.--pads 0 20 0 0
. - If you see the mouth position dislocated or some weird artifacts such as two mouths, then it can be because of over-smoothing the face detections. Use the
--nosmooth
argument and give it another try. - Experiment with the
--resize_factor
argument, to get a lower-resolution video. Why? The models are trained on faces that were at a lower resolution. You might get better, visually pleasing results for 720p videos than for 1080p videos (in many cases, the latter works well too). - The Wav2Lip model without GAN usually needs more experimenting with the above two to get the most ideal results, and sometimes, can give you a better result as well.
Our models are trained on LRS2. See here for a few suggestions regarding training on other datasets.
data_root (mvlrs_v1)
โโโ main, pretrain (we use only main folder in this work)
| โโโ list of folders
| โ โโโ five-digit numbered video IDs ending with (.mp4)
Place the LRS2 filelists (train, val, test) .txt
files in the filelists/
folder.
python preprocess.py --data_root data_root/main --preprocessed_root lrs2_preprocessed/
Additional options like batch_size
and the number of GPUs to use in parallel to use can also be set.
preprocessed_root (lrs2_preprocessed)
โโโ list of folders
| โโโ Folders with five-digit numbered video IDs
| โ โโโ *.jpg
| โ โโโ audio.wav
There are two major steps: (i) Train the expert lip-sync discriminator, (ii) Train the Wav2Lip model(s).
You can download the pre-trained weights if you want to skip this step. To train it:
python color_syncnet_train.py --data_root lrs2_preprocessed/ --checkpoint_dir <folder_to_save_checkpoints>
You can either train the model without the additional visual quality discriminator (< 1 day of training) or use the discriminator (~2 days). For the former, run:
python wav2lip_train.py --data_root lrs2_preprocessed/ --checkpoint_dir <folder_to_save_checkpoints> --syncnet_checkpoint_path <path_to_expert_disc_checkpoint>
To train with the visual quality discriminator, you should run hq_wav2lip_train.py
instead. The arguments for both files are similar. In both cases, you can resume training as well. Look at python wav2lip_train.py --help
for more details. You can also set additional less commonly-used hyper-parameters at the bottom of the hparams.py
file.
Training on other datasets might require modifications to the code. Please read the following before you raise an issue:
- You might not get good results by training/fine-tuning on a few minutes of a single speaker. This is a separate research problem, to which we do not have a solution yet. Thus, we would most likely not be able to resolve your issue.
- You must train the expert discriminator for your own dataset before training Wav2Lip.
- If it is your own dataset downloaded from the web, in most cases, needs to be sync-corrected.
- Be mindful of the FPS of the videos of your dataset. Changes to FPS would need significant code changes.
- The expert discriminator's eval loss should go down to ~0.25 and the Wav2Lip eval sync loss should go down to ~0.2 to get good results.
When raising an issue on this topic, please let us know that you are aware of all these points.
We have an HD model trained on a dataset allowing commercial usage. The size of the generated face will be 192 x 288 in our new model.
Please check the evaluation/
folder for the instructions.
This repository can only be used for personal/research/non-commercial purposes. However, for commercial requests, please contact us directly at [email protected] or [email protected]. We have an HD model trained on a dataset allowing commercial usage. The size of the generated face will be 192 x 288 in our new model. Please cite the following paper if you use this repository:
@inproceedings{10.1145/3394171.3413532,
author = {Prajwal, K R and Mukhopadhyay, Rudrabha and Namboodiri, Vinay P. and Jawahar, C.V.},
title = {A Lip Sync Expert Is All You Need for Speech to Lip Generation In the Wild},
year = {2020},
isbn = {9781450379885},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3394171.3413532},
doi = {10.1145/3394171.3413532},
booktitle = {Proceedings of the 28th ACM International Conference on Multimedia},
pages = {484โ492},
numpages = {9},
keywords = {lip sync, talking face generation, video generation},
location = {Seattle, WA, USA},
series = {MM '20}
}
Parts of the code structure are inspired by this TTS repository. We thank the author for this wonderful code. The code for Face Detection has been taken from the face_alignment repository. We thank the authors for releasing their code and models. We thank zabique for the tutorial collab notebook.