We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.IV

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Image and Video Processing

Title: AG-CUResNeSt: A Novel Method for Colon Polyp Segmentation

Abstract: Colorectal cancer is among the most common malignancies and can develop from high-risk colon polyps. Colonoscopy is an effective screening tool to detect and remove polyps, especially in the case of precancerous lesions. However, the missing rate in clinical practice is relatively high due to many factors. The procedure could benefit greatly from using AI models for automatic polyp segmentation, which provide valuable insights for improving colon polyp detection. However, precise segmentation is still challenging due to variations of polyps in size, shape, texture, and color. This paper proposes a novel neural network architecture called AG-CUResNeSt, which enhances Coupled UNets using the robust ResNeSt backbone and attention gates. The network is capable of effectively combining multi-level features to yield accurate polyp segmentation. Experimental results on five popular benchmark datasets show that our proposed method achieves state-of-the-art accuracy compared to existing methods.
Subjects: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2105.00402 [eess.IV]
  (or arXiv:2105.00402v2 [eess.IV] for this version)

Submission history

From: Sang Dinh [view email]
[v1] Sun, 2 May 2021 06:36:36 GMT (24302kb,D)
[v2] Tue, 4 May 2021 18:57:57 GMT (24255kb,D)

Link back to: arXiv, form interface, contact.