Architecture of Deep Convolutional Encoder-Decoder Networks for Building Footprint Semantic Segmentation

Figure 4

Architecture overview of PSPNet.