Fig. 4
From: Richer fusion network for breast cancer classification based on multimodal data
Schematic overview of the richer multilevel feature representation. We use average pooling on the final output of the third, fourth and fifth convolution layers of VGG16 network. Then, we concatenate the three vectors into a 1280D dimensional vector, which is used as the richer multilevel feature representation of the pathological image. The convolutional layer parameters are denoted as “Conv<number of convolutional layers>-<number of channels>”