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.
@article{arxiv.2109.01915,
title = {Sparse Spatial Attention Network for Semantic Segmentation},
author = {Mengyu Liu and Hujun Yin},
journal= {arXiv preprint arXiv:2109.01915},
year = {2021}
}