As I know, there are many algorithms to process face recognition by using recognition points. For twins recognition, we need to increase the number of recognition points, because twins also have some various features. However, this will reduce the algorithm performance. And maybe just a little change on face, the algorithm cannot recognize. If we reduce the number of recognition points, face of twins may be passed over. My question is: How to perform face recognition for twins, but guarantee can recognize human if his face has a little changes (after waking up, getting older, making up etc.)? Or any recommendations!

More Dang N. H. Thanh's questions See All
Similar questions and discussions