MPANet: Multi-Patch Attention For Infrared Small Target object Detection

TL;DR

This paper introduces MPANet, a novel neural network architecture that enhances infrared small target detection by focusing on small features and multi-scale information, outperforming existing methods in complex backgrounds.

Contribution

The paper proposes a multi-patch attention network with an axial-attention encoder and multi-scale patch branch, improving small target detection in infrared imagery.

Findings

MPANet outperforms state-of-the-art methods on the SIRST dataset.

The axial-attention encoder effectively highlights small targets.

Multi-scale patch fusion improves feature representation.

Abstract

Infrared small target detection (ISTD) has attracted widespread attention and been applied in various fields. Due to the small size of infrared targets and the noise interference from complex backgrounds, the performance of ISTD using convolutional neural networks (CNNs) is restricted. Moreover, the constriant that long-distance dependent features can not be encoded by the vanilla CNNs also impairs the robustness of capturing targets' shapes and locations in complex scenarios. To this end, a multi-patch attention network (MPANet) based on the axial-attention encoder and the multi-scale patch branch (MSPB) structure is proposed. Specially, an axial-attention-improved encoder architecture is designed to highlight the effective features of small targets and suppress background noises. Furthermore, the developed MSPB structure fuses the coarse-grained and fine-grained features from different semantic scales. Extensive experiments on the SIRST dataset show the superiority performance and effectiveness of the proposed MPANet compared to the state-of-the-art methods.