Currently I don't know which library I should use. I don't want any recommandations, I just want to know the differences, so I can decide for myself.
The available libraries are:
- OpenFace
- OpenBR
- OpenCV
My system hast to recognize about 10 different persons and it will run on a Raspberry Pi 3 with the Raspberry Pi Camera Module v2. The light conditions will differ and it the system should work with and without glasses. Furthermore it has to recognize a person in less than 2 seconds.
I know that OpenFace uses a different approach on face recognition (deep neural network) and the recognition accuracy is better. But will it work on a Raspberry Pi 3 or will it be to slow?
What are the differences between OpenCV and OpenBR? Which one has better accuracy rates? According to OpenFace OpenBR has better acceptance rates than the eigenfaces algorithm which is used by OpenCV. But OpenBR uses OpenCV and I don't know which one performes better.