Computer vision model for facial recognition. This repository comprises main tasks required to be done using facial recognition. For training run, datmo task run "python recognition.py"
1. Face Detection (detection)
- To find bounding box for faces.
- To find number of people in any image.
- Usage: Application can range like Facebook using this to find count of people and detect face to let users tag it.
2. Facial Landmark Detection (landmark_detection)
- To find the face feature locations (eyes, nose, etc) for each face in the image
- Usage: Application can lie to apply makeup application or is the founding basis to overlay structure as Snapchat does for lenses
3. Facial Verification (verification)
- Given an image of face, we can compare if any new image contains the same person.
- Usage: Widely used application eg. Uber uses to check if driver is the same person as registered on app
4. Facial Recognition (recognition/training)
-
In a given number of classes does any new image of a face lie in one of the classes.
-
Usage: Application lie to identify any celebrety and also used by Facebook to identify friends in moments or by Google in their Photos app.
Built using dlib and face_recognition