Firstly build docker container, then run it. Build it: sh ./build.sh Run it: sh ./run.sh

Inside the container go to /textask dir: cd /textask

Inside FaceDetection dir type "python setup.py install" Inside pytorch-insightface type "python setup.py install"

Then in /textask/pipe dir type "python3 main.py --img path-to-img"

Example of preprocessing alignment: base image:

Faces: before and after alignment