Superpixel Segmentation With Edge Guided Local-Global Attention Network
Published in IEEE Transactions on Circuits and Systems for Video Technology, 2025
Abstract: This paper introduces a novel superpixel segmentation method that leverages edge guided local-global attention networks. The approach enhances feature extraction capabilities and improves semantic understanding through advanced convolution techniques and iterative methods. The work contributes to the fields of image edge detection, object detection, and data visualization.
Authors: Mingzhu Xu, Zhengyu Sun, Yijun Hu, Haoyu Tang, Yupeng Hu, Xuemeng Song, and Liqiang Nie
DOI: 10.1109/TCSVT.2025.3587485