Interlaced Sparse Self-Attention for Semantic Segmentation

TL;DR

This paper introduces an interlaced sparse self-attention method that efficiently captures long- and short-range dependencies in semantic segmentation, reducing computational costs while maintaining high accuracy.

Contribution

It proposes a novel factorization of the self-attention mechanism into two sparse affinity matrices for improved efficiency in semantic segmentation.

Findings

Reduces computation and memory complexity significantly.

Achieves competitive results on six semantic segmentation benchmarks.

Effectively models both long- and short-range dependencies.

Abstract

In this paper, we present a so-called interlaced sparse self-attention approach to improve the efficiency of the \emph{self-attention} mechanism for semantic segmentation. The main idea is that we factorize the dense affinity matrix as the product of two sparse affinity matrices. There are two successive attention modules each estimating a sparse affinity matrix. The first attention module is used to estimate the affinities within a subset of positions that have long spatial interval distances and the second attention module is used to estimate the affinities within a subset of positions that have short spatial interval distances. These two attention modules are designed so that each position is able to receive the information from all the other positions. In contrast to the original self-attention module, our approach decreases the computation and memory complexity substantially especially when processing high-resolution feature maps. We empirically verify the effectiveness of our approach on six challenging semantic segmentation benchmarks.