Sparse Focus Network for Multi-Source Remote Sensing Data Classification

TL;DR

This paper introduces a sparse focus network that employs sparse attention and cross-attention mechanisms within a Transformer framework to improve multi-source remote sensing data classification, effectively reducing irrelevant information interference.

Contribution

The proposed SF-Net innovatively integrates sparse and cross-attention in a Transformer architecture for enhanced multi-source data classification.

Findings

Outperforms existing state-of-the-art methods on Berlin and Houston2018 datasets.

Effectively reduces irrelevant information interference during feature extraction.

Enhances multi-source feature interaction and fusion efficiency.

Abstract

Multi-source remote sensing data classification has emerged as a prominent research topic with the advancement of various sensors. Existing multi-source data classification methods are susceptible to irrelevant information interference during multi-source feature extraction and fusion. To solve this issue, we propose a sparse focus network for multi-source data classification. Sparse attention is employed in Transformer block for HSI and SAR/LiDAR feature extraction, thereby the most useful self-attention values are maintained for better feature aggregation. Furthermore, cross-attention is used to enhance multi-source feature interactions, and further improves the efficiency of cross-modal feature fusion. Experimental results on the Berlin and Houston2018 datasets highlight the effectiveness of SF-Net, outperforming existing state-of-the-art methods.