Sparse Spatial Attention Network for Semantic Segmentation

2021/09/04 by Mengyu Liu, Liu, Mengyu, Hujun Yin +1
Computer Science · Engineering · #Advanced Image and Video Retrieval Techniques #Advanced Neural Network Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Remote-Sensing Image Classification

paper · pdf · doi:10.48550/arxiv.2109.01915

openalex publication_date 2021/09/04 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

The spatial attention mechanism captures long-range dependencies by aggregating global contextual information to each query location, which is beneficial for semantic segmentation. In this paper, we present a sparse spatial attention network (SSANet) to improve the efficiency of the spatial attention mechanism without sacrificing the performance. Specifically, a sparse non-local (SNL) block is proposed to sample a subset of key and value elements for each query element to capture long-range relations adaptively and generate a sparse affinity matrix to aggregate contextual information efficiently. Experimental results show that the proposed approach outperforms other context aggregation methods and achieves state-of-the-art performance on the Cityscapes, PASCAL Context and ADE20K datasets.

Citations

• DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
• Semantic Understanding of Scenes Through the ADE20K Dataset
• Rethinking Atrous Convolution for Semantic Image Segmentation

Related

• Fully Attentional Network for Semantic Segmentation
• GSANet: Semantic Segmentation with Global and Selective Attention
• Cross Attention Network for Semantic Segmentation
• AASeg: Attention Aware Network for Real Time Semantic Segmentation
• Interlaced Sparse Self-Attention for Semantic Segmentation