Open YakirHasid opened 8 hours ago
Currently the backend of Offline classifies using the whole image and not only the face, it's incorrect. The correct way is to send the model the image of only the face itself.
Currently the backend of Offline classifies using the whole image and not only the face, it's incorrect. The correct way is to send the model the image of only the face itself.