OeslleLucena / FASNet

CNN archictecture for face anti-spoofing
MIT License
177 stars 56 forks source link

the model cross-validation results are not satisfactory? #7

Open hanson-young opened 6 years ago

hanson-young commented 6 years ago

I am working on face anti-spoofing attack recently. I have tried lots of methods, including LBP, HOG + SVM, CNN and so on. However, I found that the models trained on casia dataset and replay attack dataset are not generalized very good, cross-validation results are not satisfactory, I also try to use your per-train VGG model, also have the same problem

lotfiGipsyKing commented 6 years ago

same thing for me, i tested the pretrained VGG model, and is not accurate