L*ReLU: Piece-wise Linear Activation Functions for Deep Fine-grained Visual Categorization

Open in new window