Selective Structured State Space for Multispectral-fused Small Target Detection

TL;DR

This paper introduces a novel approach combining enhanced local attention modules and multispectral fusion techniques to improve small target detection accuracy in high-resolution remote sensing imagery, while maintaining computational efficiency.

Contribution

It develops the ESTD and CARG modules to enhance small target detection and proposes the MEPF module for effective multispectral feature fusion, addressing limitations of existing models.

Findings

Improved detection accuracy for small targets.

Reduced computational complexity with linear attention mechanisms.

Effective multispectral feature fusion enhances target visibility.

Abstract

Target detection in high-resolution remote sensing imagery faces challenges due to the low recognition accuracy of small targets and high computational costs. The computational complexity of the Transformer architecture increases quadratically with image resolution, while Convolutional Neural Networks (CNN) architectures are forced to stack deeper convolutional layers to expand their receptive fields, leading to an explosive growth in computational demands. To address these computational constraints, we leverage Mamba's linear complexity for efficiency. However, Mamba's performance declines for small targets, primarily because small targets occupy a limited area in the image and have limited semantic information. Accurate identification of these small targets necessitates not only Mamba's global attention capabilities but also the precise capture of fine local details. To this end, we enhance Mamba by developing the Enhanced Small Target Detection (ESTD) module and the Convolutional Attention Residual Gate (CARG) module. The ESTD module bolsters local attention to capture fine-grained details, while the CARG module, built upon Mamba, emphasizes spatial and channel-wise information, collectively improving the model's ability to capture distinctive representations of small targets. Additionally, to highlight the semantic representation of small targets, we design a Mask Enhanced Pixel-level Fusion (MEPF) module for multispectral fusion, which enhances target features by effectively fusing visible and infrared multimodal information.