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

Figure 8

An example of the original and ground truth mask of testing Massachusetts buildings’ dataset.