This repository uses code from first order model and openCV to generate alive photos.
pip3 install -r requirements.txt
Download pretrained model here.. You only need to down "vox-adv-cpk.pth.tar"
Extract model and put it in "checkpoints" dir.
python3 demo.py --config config/vox-adv-256.yaml --driving_video driving_video/short.mp4 --source_image source_image/friends.jpeg --checkpoint checkpoints/vox-adv-cpk.pth.tar --relative --adapt_scale
- Apply multiple faces reenactment