[Retracted] DTFA-Net: Dynamic and Texture Features Fusion Attention Network for Face Antispoofing

Table 2

The structure of fusion attention module.

Layer	Input size	Kernel size	Filter	Stride
GlobalAvgPool	13  13  512	13  13		1
Fc1	512
Fc2	32
LK_Conv	13  13  512	3  3  512	256	2
LK_MaxPool	6  6  256	3  3		1