Skip to main content
Fig. 7 | EURASIP Journal on Image and Video Processing

Fig. 7

From: Deep-learned faces: a survey

Fig. 7

DeepID2+ net. Conv n, nth convolutional layer (with max-pooling). FC n, nth fully connected layer. Id, identification supervisory signal. Ve, verification supervisory signal. Dashed arrows denote forward propagation. Solid arrows denote supervisory signals. Nets in the left and right are two DeepID2+ nets with shared weights and different input faces

Back to article page