Deep-Live-Cam

Real-time face swap and video deepfake with a single click and only a single image.

Disclaimer

This software is intended as a productive contribution to the AI-generated media industry. It aims to assist artists with tasks like animating custom characters or using them as models for clothing, etc.

We are aware of the potential for unethical applications and are committed to preventative measures. A built-in check prevents the program from processing inappropriate media (nudity, graphic content, sensitive material like war footage, etc.). We will continue to develop this project responsibly, adhering to law and ethics. We may shut down the project or add watermarks if legally required.

Users are expected to use this software responsibly and legally. If using a real person’s face, obtain their consent and clearly label any output as a deepfake when sharing online. We are not responsible for end-user actions.

Quick Start (Windows / Nvidia)

Download latest pre-built version with CUDA support - No Manual Installation/Downloading required.

Installation (Manual)

Please be aware that the installation needs technical skills and is NOT for beginners, consider downloading the prebuilt. Please do NOT open platform and installation related issues on GitHub before discussing it on the discord server.

Basic Installation (CPU)

This is more likely to work on your computer but will be slower as it utilizes the CPU.

1. Setup Your Platform

• Python (3.10 recommended)
• pip
• git
• ffmpeg
• Visual Studio 2022 Runtimes (Windows)

2. Clone Repository

https://github.com/hacksider/Deep-Live-Cam.git

3. Download Models

1. GFPGANv1.4
2. inswapper_128_fp16.onnx (Note: Use this replacement version if you encounter issues)

Place these files in the “models” folder.

4. Install Dependencies

We highly recommend using a venv to avoid issues.

pip install -r requirements.txt

For macOS: Install or upgrade the python-tk package:

brew install [email protected]

Run: If you don’t have a GPU, you can run Deep-Live-Cam using python run.py. Note that initial execution will download models (~300MB).

GPU Acceleration (Optional)

pip uninstall onnxruntime onnxruntime-gpu
pip install onnxruntime-gpu==1.16.3

python run.py --execution-provider cuda

pip uninstall onnxruntime onnxruntime-silicon
pip install onnxruntime-silicon==1.13.1

python run.py --execution-provider coreml

pip uninstall onnxruntime onnxruntime-coreml
pip install onnxruntime-coreml==1.13.1

python run.py --execution-provider coreml

pip uninstall onnxruntime onnxruntime-directml
pip install onnxruntime-directml==1.15.1

python run.py --execution-provider directml

pip uninstall onnxruntime onnxruntime-openvino
pip install onnxruntime-openvino==1.15.0

python run.py --execution-provider openvino

Usage

1. Image/Video Mode

• Execute python run.py.
• Choose a source face image and a target image/video.
• Click “Start”.
• The output will be saved in a directory named after the target video.

2. Webcam Mode

• Execute python run.py.
• Select a source face image.
• Click “Live”.
• Wait for the preview to appear (10-30 seconds).
• Use a screen capture tool like OBS to stream.
• To change the face, select a new source image.

Features

Resizable Preview Window

Dynamically improve performance using the --live-resizable parameter.

Face Mapping

Track and change faces on the fly.

Source Video:

Enable Face Mapping:

Map the Faces:

See the Magic!

Watch movies in realtime:

It’s as simple as opening a movie on the screen, and selecting OBS as your camera!

Benchmarks

On Deepware scanner - Most popular deepfake detection website, recording of realtime faceswap ran on an RTX 3060 -

Command Line Arguments

options:
  -h, --help                                               show this help message and exit
  -s SOURCE_PATH, --source SOURCE_PATH                     select a source image
  -t TARGET_PATH, --target TARGET_PATH                     select a target image or video
  -o OUTPUT_PATH, --output OUTPUT_PATH                     select output file or directory
  --frame-processor FRAME_PROCESSOR [FRAME_PROCESSOR ...]  frame processors (choices: face_swapper, face_enhancer, ...)
  --keep-fps                                               keep original fps
  --keep-audio                                             keep original audio
  --keep-frames                                            keep temporary frames
  --many-faces                                             process every face
  --map-faces                                              map source target faces
  --nsfw-filter                                            filter the NSFW image or video
  --video-encoder {libx264,libx265,libvpx-vp9}             adjust output video encoder
  --video-quality [0-51]                                   adjust output video quality
  --live-mirror                                            the live camera display as you see it in the front-facing camera frame
  --live-resizable                                         the live camera frame is resizable
  --max-memory MAX_MEMORY                                  maximum amount of RAM in GB
  --execution-provider {cpu} [{cpu} ...]                   available execution provider (choices: cpu, ...)
  --execution-threads EXECUTION_THREADS                    number of execution threads
  -v, --version                                            show program's version number and exit

Looking for a CLI mode? Using the -s/–source argument will make the run program in cli mode.

Webcam Mode on WSL2 Ubuntu (Optional)

usbipd list  
usbipd bind --busid x-x # Replace x-x with your webcam's bus ID  
usbipd attach --wsl --busid x-x # Replace x-x with your webcam's bus ID  

git clone [https://github.com/hacksider/Deep-Live-Cam](https://github.com/hacksider/Deep-Live-Cam)  

zcat /proc/config.gz | grep -i "CONFIG_USB_VIDEO_CLASS"  

ls /lib/modules/$(uname -r)/kernel/drivers/media/usb/uvc/  

sudo modprobe uvcvideo  
dmesg | tail  

sudo ls -al /dev/video*  

sudo usermod -a -G video $USER  
sudo chgrp video /dev/video0 /dev/video1  
sudo chmod 660 /dev/video0 /dev/video1  

sudo nano /etc/udev/rules.d/81-webcam.rules  

KERNEL=="video[0-9]*", GROUP="video", MODE="0660"  

sudo udevadm control --reload-rules && sudo udevadm trigger  

[ONNXRuntimeError] : 1 : FAIL : Failed to load library [libonnxruntime_providers_cuda.so](http://libonnxruntime_providers_cuda.so/) with error: libcufft.so.10: cannot open shared object file: No such file or directory  

/usr/local/cuda/bin/nvcc --version  

sudo apt-get -y install cuda-toolkit-11-8  

Future Updates & Roadmap

For the latest experimental builds and features, see the experimental branch.

TODO:

• [ ] Develop a version for web app/service
• [ ] Speed up model loading
• [ ] Speed up real-time face swapping
• [x] Support multiple faces
• [x] UI/UX enhancements for desktop app

This is an open-source project developed in our free time. Updates may be delayed.

Tips and Links:

• How to make the most of Deep-Live-Cam
• Face enhancer is good, but still very slow for any live streaming purpose.

Credits

• ffmpeg: for making video related operations easy
• deepinsight: for their insightface project which provided a well-made library and models. Please be reminded that the use of the model is for non-commercial research purposes only.
• havok2-htwo : for sharing the code for webcam
• GosuDRM : for open version of roop
• pereiraroland26 : Multiple faces support
• vic4key : For supporting/contributing on this project
• KRSHH : For his contributions
• and all developers behind libraries used in this project.
• Foot Note: Please be informed that the base author of the code is s0md3v

Contributions

Star History