Token Clustering and Semantic Sequence Mamba for Hyperspectral Image Classification

2026-09-23

AI Summary

This study addresses the challenge of spectral-spatial heterogeneity in hyperspectral image classification and the limitation of existing Mamba models that overlook semantic similarity. To this end, we propose STMamba, which innovatively incorporates density-aware clustering and a quadtree-based dynamic selection mechanism to construct semantically coherent sparse token sequences. Furthermore, a parallel SWSM module is designed to efficiently capture long-range dependencies while suppressing irrelevant interactions, integrated within a hierarchical encoder-decoder architecture for precise feature extraction. Extensive experiments on three large-scale benchmark datasets demonstrate that STMamba significantly outperforms current state-of-the-art methods in both classification accuracy and robustness.

Abstract

Although hyperspectral images (HSIs) provide rich spectral-spatial information, accurate pixel-level classification remains challenging because of spectral-spatial heterogeneity and complex spatial structures. Existing vision state space models (Mamba) typically construct sequences according to predefined spatial neighborhoods, without explicitly accounting for semantic similarity or spatial non-stationarity. To address this limitation, we propose Token Clustering and Semantic Sequence Mamba (STMamba), which organizes sparse tokens into semantically coherent sequences for hyperspectral image classification with the following features. First, at the macro level, a hierarchical encoder decoder progressively selects semantic tokens with the Token Clustering Module (TCM) and restores dense features using a parameter-free Cross-scale Neighborhood Attention (CNA) Upsampler. Second, at the micro level, TCM first identifies representative cluster centers through density-aware clustering and estimates soft memberships based on feature similarity. A quadtree-based dynamic selection strategy then retains sparse and spatially distributed tokens from each semantic cluster, forming coherent semantic-token sequences while reducing redundant pixel-wise representations. Third, parallel Spatial and Spectral Semantic-wise Sequencing Mamba (SWSM) modules capture complementary long-range spatial and spectral dependencies within homogeneous semantic token sequences while suppressing irrelevant interactions across heterogeneous regions. Experimental results on three large-scale benchmark datasets demonstrate that STMamba outperforms the SOTA methods with respect to quantitative and qualitative results.

Problem

• Hyperspectral Image Classification
• Vision State Space Model
• Semantic Similarity
• Spatial Non-stationarity
• Spectral-Spatial Heterogeneity

Innovation

• Hyperspectral Image Classification
• Mamba
• Token Clustering
• Semantic Sequence
• State Space Model