SwinCup: Cascaded swin transformer for histopathological structures segmentation in colorectal cancer

Usama Zidan, Mohamed Medhat Gaber, Mohammed M. Abdelsamea

Abstract: Highlights•Hierarchical transformer-based models can extract global context features.•Cascaded up samplers to extract features without consuming much computational cost.•Globally aware models generate better segmentation results in microscopy images.•In contrastive self-supervised learning, adding labels improves results.