Gated Path Selection Network for Semantic Segmentation

• Published in: IEEE transactions on image processing Vol. 30; pp. 2436 - 2449
• Main Authors: Geng, Qichuan, Zhang, Hong, Qi, Xiaojuan, Huang, Gao, Yang, Ruigang, Zhou, Zhong
• Format: Journal Article
• Language: English
• Published: United States IEEE 2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: adaptive context aggregation; adaptive receptive fields and sampling locations; Adaptive sampling; Aggregates; Bells; Context; Context modeling; Convolution; Free form; Global Positioning System; Image segmentation; local discriminative feature; Logic gates; Semantic segmentation; Semantics
• ISSN: 1057-7149; 1941-0042; 1941-0042
• DOI: 10.1109/TIP.2020.3046921

Semantic segmentation is a challenging task that needs to handle large scale variations, deformations, and different viewpoints. In this paper, we develop a novel network named Gated Path Selection Network (GPSNet), which aims to adaptively select receptive fields while maintaining the dense sampling capability. In GPSNet, we first design a two-dimensional SuperNet, which densely incorporates features from growing receptive fields. And then, a Comparative Feature Aggregation (CFA) module is introduced to dynamically aggregate discriminative semantic context. In contrast to previous works that focus on optimizing sparse sampling locations on regular grids, GPSNet can adaptively harvest free form dense semantic context information. The derived adaptive receptive fields and dense sampling locations are data-dependent and flexible which can model various contexts of objects. On two representative semantic segmentation datasets, i.e ., Cityscapes and ADE20K, we show that the proposed approach consistently outperforms previous methods without bells and whistles.