DSXFormer: Dual-Pooling Spectral Squeeze-Expansion and Dynamic Context Attention Transformer for Hyperspectral Image Classification

2026-02-02

AI Summary

This work addresses the challenges of hyperspectral image classification, including high spectral dimensionality, complex spectral-spatial correlations, and limited labeled samples, which hinder existing Transformer models from balancing spectral discriminability and computational efficiency. To this end, we propose DSXFormer, which introduces a novel Dual-pooling Spectral Compression-Expansion (DSX) module that integrates global average and max pooling to enhance spectral channel recalibration. Additionally, a Dynamic Contextual Attention (DCA) mechanism is designed to efficiently model local spectral-spatial dependencies within a windowed Transformer framework while reducing computational overhead. Coupled with a multi-scale patch extraction and merging strategy, DSXFormer achieves state-of-the-art classification accuracies of 99.95%, 98.91%, 99.85%, and 98.52% on the Salinas, Indian Pines, Pavia University, and Kennedy Space Center benchmark datasets, respectively, significantly outperforming existing methods.

Technology Category

• Machine Learning: Feature Construction/Reformulation
• Computer Vision: Multi-modal Vision
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Hyperspectral image classification (HSIC) is a challenging task due to high spectral dimensionality, complex spectral-spatial correlations, and limited labeled training samples. Although transformer-based models have shown strong potential for HSIC, existing approaches often struggle to achieve sufficient spectral discriminability while maintaining computational efficiency. To address these limitations, we propose a novel DSXFormer, a novel dual-pooling spectral squeeze-expansion transformer with Dynamic Context Attention for HSIC. The proposed DSXFormer introduces a Dual-Pooling Spectral Squeeze-Expansion (DSX) block, which exploits complementary global average and max pooling to adaptively recalibrate spectral feature channels, thereby enhancing spectral discriminability and inter-band dependency modeling. In addition, DSXFormer incorporates a Dynamic Context Attention (DCA) mechanism within a window-based transformer architecture to dynamically capture local spectral-spatial relationships while significantly reducing computational overhead. The joint integration of spectral dual-pooling squeeze-expansion and DCA enables DSXFormer to achieve an effective balance between spectral emphasis and spatial contextual representation. Furthermore, patch extraction, embedding, and patch merging strategies are employed to facilitate efficient multi-scale feature learning. Extensive experiments conducted on four widely used hyperspectral benchmark datasets, including Salinas (SA), Indian Pines (IP), Pavia University (PU), and Kennedy Space Center (KSC), demonstrate that DSXFormer consistently outperforms state-of-the-art methods, achieving classification accuracies of 99.95%, 98.91%, 99.85%, and 98.52%, respectively.

Problem

• Hyperspectral image classification
• spectral discriminability
• spectral-spatial correlations
• limited labeled samples
• computational efficiency

Innovation

• Dual-Pooling Spectral Squeeze-Expansion
• Dynamic Context Attention
• Hyperspectral Image Classification
• Transformer Architecture
• Spectral-Spatial Feature Learning