Simbotic Faceswap

Docker container for Faceswap python application.

Prerequisites

Docker CE 18.x
docker-compose 1.23.x
Nvidia Docker

Docker

Please refer to Offical docker documentation to install docker for Ubuntu 18.04

Docker Compose

Official docker compose installation

Nvidia Docker

Official repository

# Add the package repositories

curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | \
 sudo apt-key add -
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | \
 sudo tee /etc/apt/sources.list.d/nvidia-docker.list
sudo apt-get update


# Install nvidia-docker2 and reload the Docker daemon configuration
sudo apt-get install -y nvidia-docker2
sudo pkill -SIGHUP dockerd

Allow X Server connections

For open windows/guis inside the container, in your host machine type the following:

xhost +

Now we are ready to proceed

Build and Running

Clone the repo

git clone git@github.com:VertexStudio/simbotic-faceswap.git

Execute the setup.sh

./setup.sh

Inside the recent clone repo type:

docker-compose up -d --build

This will build and start the container.

Test the installation

./test_installation.sh

It should print, everythin related with tensorflow-gpu

Running

Execute the gui from faceswap using the bash script:

./run_gui.sh

Command Line usage

Extract faces

Run the following script

./extract_photos.sh /data/path/to/photos /data/path/to/faces

Where path/to/photos should reside inside data directory created with ./setup.sh script and it would contain all the photos that we need to extract the faces.

path/to/faces should reside inside data directory created with ./setup.sh script it would contain all the extracted faces after the process finish.

Example:

./extract_photos.sh /data/source/yoda /data/faces/yoda

Train model

Run the script

./train_model.sh /data/path/to/A/faces /data/path/to/B/faces /data/path/to/modelAB

Same as the Extract faces part, but in this case we need to specify where the model should be saved (following the /data/ as a parent directory)

./train_model.sh /data/faces/liam_neeson /data/faces/yoda /data/model/liam_yoda_model/

Troubleshooting

Tensorflow raised an unknown error. This is most likely caused by a failure to launch cuDNN which can occur for some GPU/Tensorflow combinations. You should enable `allow_growth` to attempt to resolve this issue

To solve this issue, you need to change allow_growth to True inside faceswap/config/extract.ini