The Chained Context Aggregation Network (CANet) is a special image segmentation network that employs an asymmetric decoder to recover precise spatial details of prediction maps.