anti spoofing failed at a print face

[tomjerrygithub]

is there any better solution to do this ?

[syaringan357]

I just updated the model today. You can try it. I will update the final model next week. My computer training is too slow, so I only used 8 convolutional kernels, and I've only trained 14 epochs so far. But Dr. Liu Xiaojie's source code used 32 convolutional kernels, which needs 40 epochs of training by default. If you have the conditions you can train yourself, the source code and data is here. https://github.com/yaojieliu/CVPR2019-DeepTreeLearningForZeroShotFaceAntispoofing http://cvlab.cse.msu.edu/project-face-anti.html

[tomjerrygithub]

thanks i will try