Graph-Segmenter:graph transformer with boundary-aware attention for semantic segmentation  

作　　者：Zizhang WU Yuanzhu GAN Tianhao XU Fan WANG 

机构地区：[1]Computer Vision Perception Department of ZongMu Technology,Shanghai 201203,China [2]Faculty of Electrical Engineering,Information Technology,Physics,Technical University of Braunschweig,Braunschweig 38106,Germany

出　　处：《Frontiers of Computer Science》2024年第5期97-108,共12页计算机科学前沿（英文版）

摘　　要：Thetransformer-based semantic segmentation approaches,which divide the image into different regions by sliding windows and model the relation inside each window,have achieved outstanding success.However,since the relation modeling between windows was not the primary emphasis of previous work,it was not fully utilized.To address this issue,we propose a Graph-Segmenter,including a graph transformer and a boundary-aware attention module,which is an effective network for simultaneously modeling the more profound relation between windows in a global view and various pixels inside each window as a local one,and for substantial low-cost boundary adjustment.Specifically,we treat every window and pixel inside the window as nodes to construct graphs for both views and devise the graph transformer.The introduced boundary-awareattentionmoduleoptimizes theedge information of the target objects by modeling the relationship between the pixel on the object's edge.Extensive experiments on three widely used semantic segmentation datasets(Cityscapes,ADE-20k and PASCAL Context)demonstrate that our proposed network,a Graph Transformer with Boundary-aware Attention,can achieve state-of-the-art segmentation performance.

关 键 词：graph transformer graph relation network boundary-aware ATTENTION semantic segmentation 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]