Open sagar1garg opened 5 years ago
Hi,
Performance is directly dependent on facial landmark detection. The landmark detector used here (the one from dlib) requires face detection in every frame which really slows things down. If you want it to work faster, you would probably have to use a different landmark detector. There are many available on GitHub: Example 1, my work Example 2, someone else's work
This method does not use any training.
Best regards,
Marek
@MarekKowalski will this work in real time scenarios for both pictures and short videos? What would be the training time requirements for the model to generate optimum results?