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

Figure 5

An illustration of the LinkNet architecture.