gcDLSeg: Integrating Graph-cut into Deep Learning for Binary Semantic Segmentation

Hui Xie; Weiyu Xu; Ya Xing Wang; John Buatti; Xiaodong Wu

doi:10.48550/arxiv.2312.04713

Back

gcDLSeg: Integrating Graph-cut into Deep Learning for Binary Semantic Segmentation

Preprint

Open access

gcDLSeg: Integrating Graph-cut into Deep Learning for Binary Semantic Segmentation

Hui Xie, Weiyu Xu, Ya Xing Wang, John Buatti and Xiaodong Wu

arXiv.org

Cornell University

12/07/2023

DOI: 10.48550/arxiv.2312.04713

Files and links (1)

url

https://doi.org/10.48550/arxiv.2312.04713View

Preprint (Author's original)This preprint has not been evaluated by subject experts through peer review. Preprints may undergo extensive changes and/or become peer-reviewed journal articles. Open Access

Abstract

Binary semantic segmentation in computer vision is a fundamental problem. As a model-based segmentation method, the graph-cut approach was one of the most successful binary segmentation methods thanks to its global optimality guarantee of the solutions and its practical polynomial-time complexity. Recently, many deep learning (DL) based methods have been developed for this task and yielded remarkable performance, resulting in a paradigm shift in this field. To combine the strengths of both approaches, we propose in this study to integrate the graph-cut approach into a deep learning network for end-to-end learning. Unfortunately, backward propagation through the graph-cut module in the DL network is challenging due to the combinatorial nature of the graph-cut algorithm. To tackle this challenge, we propose a novel residual graph-cut loss and a quasi-residual connection, enabling the backward propagation of the gradients of the residual graph-cut loss for effective feature learning guided by the graph-cut segmentation model. In the inference phase, globally optimal segmentation is achieved with respect to the graph-cut energy defined on the optimized image features learned from DL networks. Experiments on the public AZH chronic wound data set and the pancreas cancer data set from the medical segmentation decathlon (MSD) demonstrated promising segmentation accuracy, and improved robustness against adversarial attacks.

Computer Science - Artificial Intelligence

Computer Science - Computer Vision and Pattern Recognition

Computer Science - Learning

Details

Title: Subtitle: gcDLSeg: Integrating Graph-cut into Deep Learning for Binary Semantic Segmentation
Creators: Hui Xie
Weiyu Xu - University of Iowa
Ya Xing Wang - Beijing Tongren Hospital
John Buatti - University of Iowa
Xiaodong Wu
Resource Type: Preprint
Publication Details: arXiv.org
DOI: 10.48550/arxiv.2312.04713
eISSN: 2331-8422
Publisher: Cornell University; Ithaca, New York
Language: English
Date posted: 12/07/2023
Academic Unit: Electrical and Computer Engineering; Radiation Oncology; The Iowa Institute for Biomedical Imaging; Neurosurgery; Otolaryngology
Record Identifier: 9984528115702771

Metrics

21 Record Views