Deep co-supervision and attention fusion strategy for automatic COVID-19 lung infection segmentation on CT images

Highlights：

• An ESM is presented to highlight low-level boundary features, and the edge supervised information is incorporated into the initial stage of down-sampling.

• An ASSM is proposed to enhance high-level semantics from feature maps with different scales, and the mask supervised information is introduced into the later stage of down-sampling.

• An AFM is developed to fuse various scale feature maps from the up-sampling stage. An attention mechanism is utilized to reduce the semantic gaps between high-level and low-level feature maps, so as to strengthen and supplement the lost detailed information in high-level representations.

• A joint loss function is constructed by combining the edge supervised loss, auxiliary semantic supervised loss and fusion loss, thereby achieving a deep collaborative supervision on edges and semantics.