matteliot/roop

Take a video and replace the face in it with a face of your choice. You only need one image of the desired face. No dataset, no training.

That's it, that's the software. You can watch some demos here.

Installation

Do not create any issues regarding installation problems. I am only responsible for issues in this program, use google for help.

install python, pip and git
install ffmpeg
run the following commands in terminal:

git clone https://github.com/s0md3v/roop
cd roop
pip install -r requirements.txt

Download this file and keep it in roop directory. Mirror #1, Mirror #2, Mirror #3
If you are on windows, you will need to install Microsoft Visual C++ 14.0. Google how to do that.

GPU Accleration (Optional)

If you have a good enough GPU, you can use it to speed-up the face-swapping process by running run.py with --gpu flag. If you plan on doing it, you will need to install the appropriate onnxruntime-* package as follows:

NVIDIA

Install cuda and then,

pip install onnxruntime-gpu

AMD

Install ROCM-based torch packages from here and then,

git clone https://github.com/microsoft/onnxruntime
cd onnxruntime
./build.sh --config Release --build_wheel --update --build --parallel --cmake_extra_defines CMAKE_PREFIX_PATH=/opt/rocm/lib/cmake ONNXRUNTIME_VERSION=$ONNXRUNTIME_VERSION onnxruntime_BUILD_UNIT_TESTS=off --use_rocm --rocm_home=/opt/rocm
pip install build/Linux/Release/dist/*.whl

Usage

Note: When you run this program for the first time, it will download some models ~300MB in size.

Executing python run.py command will launch this window:

Choose a face (image with desired face) and the target image/video (image/video in which you want to replace the face) and click on Start. Open file explorer and navigate to the directory you select your output to be in. You will find a directory named <video_title> where you can see the frames being swapped in realtime. Once the processing is done, it will create the output file. That's it.

Don't touch the FPS checkbox unless you know what you are doing.

Additional command line arguments are given below:

-h, --help            show this help message and exit
-f SOURCE_IMG, --face SOURCE_IMG
                        use this face
-t TARGET_PATH, --target TARGET_PATH
                        replace this face
-o OUTPUT_FILE, --output OUTPUT_FILE
                      save output to this file
--keep-fps            keep original fps
--gpu                 use gpu
--keep-frames         don't delete frames directory

Looking for a CLI mode? Using the -f/--face argument will make the program in cli mode.

Future plans

Replace a selective face throughout the video
Support for replacing multiple faces

Credits

ffmpeg: for making video related operations easy
deepinsight: for their insightface project which provided a well-made library and models.
and all developers behind libraries used in this project.

matteliot / roop