Face recognition module for deepfakes video project.

The task is to develop a recognition module for the deep fakes video project.
Solution – to make this possible, we created a 128d vector, which is a face descriptor. Then, as an intermediate result, we were able to extract it from any person in the video, and then compare it with the vectors we have. The result is obvious.

Team — 1 ML dev
Terms — 4 weeks

Have a question?

Leave a request and we will call you back.