Fig. 3From: The visual human face super-resolution reconstruction algorithm based on improved deep residual networkTraditional residual unit and improved residual unitBack to article page