search

juan-csv / face_recognition_occlusion

Face recognition implementation is capable of recognizing faces with occlusion, this includes faces wearing masks.
MIT License
80 stars 26 forks source link

about network weight #3

Open Xu-Hardy opened 3 years ago

Xu-Hardy commented 3 years ago

hi~ How can I train the nerual network in the project? Which dataset and pre-trained model did you use?

juan-csv commented 3 years ago

Hello, this is a model created by OpenCV, I leave this url for you to review the model and in this url they explain how to train the model for face detection.

To do easy recognition you need a feature vector that encodes the facial features, the easiest way is to use the face recognition library.

Xu-Hardy commented 3 years ago

I mean the caffenet you used.What did it do?

NCUTzyx commented 2 years ago

Boost.Python.ArgumentError: Python argument types in rectangle.init(rectangle, numpy.int32, numpy.int32, numpy.int32, numpy.int32) did not match C++ signature: init(struct _object ptr64, long left, long top, long right, long bottom) init__(struct _object __ptr64)

What's the problem with that?