FeDNet: Feature Decoupled Network for polyp segmentation from endoscopy images

Yanzhou Su, Jian Cheng, Chuqiao Zhong, Yijie Zhang, Jin Ye, Junjun He, Jun Liu

Abstract

Highlights•We propose a feature decoupled network (FeDNet) for polyp segmentation.•A feature decoupled module is used to decouple feature into low- and high-frequency component.•We optimize the body and edge part simultaneously.•FeDNet achieves the state-of-the-art performance in five polyp segmentation benchmarks.
BibTeX
@article{su2023fednet,
  author = {Yanzhou Su and Jian Cheng and Chuqiao Zhong and Yijie Zhang and Jin Ye and Junjun He and Jun Liu},
  title = {{FeDNet: Feature Decoupled Network for polyp segmentation from endoscopy images}},
  journal = {Biomedical Signal Processing and Control},
  year = {2023},
  url = {https://openreview.net/forum?id=tBePatkbNn},
}

Loading related content...

Figure