HLSK-CASMamba: hybrid large selective kernel and convolutional additive self-attention mamba for hyperspectral image classification
Abstract Classifying hyperspectral images (HSIs) is a key challenge in remote sensing, with convolutional neural networks (CNNs) and transformer models becoming leading techniques in this area. CNNs, while effective, often struggle to adequately capture intricate semantic features, and increasing ne...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2025-06-01
|
| Series: | Journal of King Saud University: Computer and Information Sciences |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s44443-025-00060-z |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849235358116478976 |
|---|---|
| author | Xiaoqing Wan Yupeng He Feng Chen Ziqi Sun Dongtao Mo |
| author_facet | Xiaoqing Wan Yupeng He Feng Chen Ziqi Sun Dongtao Mo |
| author_sort | Xiaoqing Wan |
| collection | DOAJ |
| description | Abstract Classifying hyperspectral images (HSIs) is a key challenge in remote sensing, with convolutional neural networks (CNNs) and transformer models becoming leading techniques in this area. CNNs, while effective, often struggle to adequately capture intricate semantic features, and increasing network depth leads to significantly higher computational costs. Conversely, transformers, despite their efficacy in modeling spectral-spatial dependencies, introduce significant computational overhead due to their complexity. Mamba, leveraging the state space model (SSM), presents a compelling alternative that efficiently captures long-range dependencies in HSIs while ensuring computational efficiency with linear complexity. To improve the classification performance of HSIs by simultaneously extracting rich local and global spatial-spectral features, as well as deep semantic features, while reducing the computational complexity of the model, this paper proposes an innovative hybrid large selective kernel and convolutional additive self-attention model (HLSK-CASMamba) for HSI classification. First, we design a feature extraction module that combines a 3D convolution layer, a 2D convolution layer, and a large selective kernel (LSK) network, enabling the efficient extraction of both depth-related and spatial details information from HSIs. Second, we propose a novel CASMamba model, with its core module, CAS-VSSM, combining convolutional additive self-attention (CAS) and the vision state-space sequence model (VSSM). This fusion leverages the local feature extraction of convolutions, spatial dependency modeling of self-attention, and long-range dependency handling of VSSM, enhancing the capture of both local and global context while ensuring computational efficiency. Finally, we incorporate the KANLinear module to replace the traditional linear layer, enhancing sample label acquisition. Extensive evaluations on three popular HSIs show that, under 10% training samples, the proposed method achieves 99.57% accuracy on the Houston 2013 dataset, 99.96% on the Botswana dataset, and 99.92% on the University of Pavia dataset, outperforming various existing advanced techniques. |
| format | Article |
| id | doaj-art-0acd62349f6041a2ae5fe435eb602075 |
| institution | Kabale University |
| issn | 1319-1578 2213-1248 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Springer |
| record_format | Article |
| series | Journal of King Saud University: Computer and Information Sciences |
| spelling | doaj-art-0acd62349f6041a2ae5fe435eb6020752025-08-20T04:02:49ZengSpringerJournal of King Saud University: Computer and Information Sciences1319-15782213-12482025-06-0137412010.1007/s44443-025-00060-zHLSK-CASMamba: hybrid large selective kernel and convolutional additive self-attention mamba for hyperspectral image classificationXiaoqing Wan0Yupeng He1Feng Chen2Ziqi Sun3Dongtao Mo4College of Computer Science and Technology, Hengyang Normal UniversityCollege of Computer Science and Technology, Hengyang Normal UniversityCollege of Computer Science and Technology, Hengyang Normal UniversityCollege of Computer Science and Technology, Hengyang Normal UniversityCollege of Computer Science and Technology, Hengyang Normal UniversityAbstract Classifying hyperspectral images (HSIs) is a key challenge in remote sensing, with convolutional neural networks (CNNs) and transformer models becoming leading techniques in this area. CNNs, while effective, often struggle to adequately capture intricate semantic features, and increasing network depth leads to significantly higher computational costs. Conversely, transformers, despite their efficacy in modeling spectral-spatial dependencies, introduce significant computational overhead due to their complexity. Mamba, leveraging the state space model (SSM), presents a compelling alternative that efficiently captures long-range dependencies in HSIs while ensuring computational efficiency with linear complexity. To improve the classification performance of HSIs by simultaneously extracting rich local and global spatial-spectral features, as well as deep semantic features, while reducing the computational complexity of the model, this paper proposes an innovative hybrid large selective kernel and convolutional additive self-attention model (HLSK-CASMamba) for HSI classification. First, we design a feature extraction module that combines a 3D convolution layer, a 2D convolution layer, and a large selective kernel (LSK) network, enabling the efficient extraction of both depth-related and spatial details information from HSIs. Second, we propose a novel CASMamba model, with its core module, CAS-VSSM, combining convolutional additive self-attention (CAS) and the vision state-space sequence model (VSSM). This fusion leverages the local feature extraction of convolutions, spatial dependency modeling of self-attention, and long-range dependency handling of VSSM, enhancing the capture of both local and global context while ensuring computational efficiency. Finally, we incorporate the KANLinear module to replace the traditional linear layer, enhancing sample label acquisition. Extensive evaluations on three popular HSIs show that, under 10% training samples, the proposed method achieves 99.57% accuracy on the Houston 2013 dataset, 99.96% on the Botswana dataset, and 99.92% on the University of Pavia dataset, outperforming various existing advanced techniques.https://doi.org/10.1007/s44443-025-00060-zHyperspectral imageClassificationConvolutional neural networks (CNNs)Large selective kernelConv additive self-attentionMamba |
| spellingShingle | Xiaoqing Wan Yupeng He Feng Chen Ziqi Sun Dongtao Mo HLSK-CASMamba: hybrid large selective kernel and convolutional additive self-attention mamba for hyperspectral image classification Journal of King Saud University: Computer and Information Sciences Hyperspectral image Classification Convolutional neural networks (CNNs) Large selective kernel Conv additive self-attention Mamba |
| title | HLSK-CASMamba: hybrid large selective kernel and convolutional additive self-attention mamba for hyperspectral image classification |
| title_full | HLSK-CASMamba: hybrid large selective kernel and convolutional additive self-attention mamba for hyperspectral image classification |
| title_fullStr | HLSK-CASMamba: hybrid large selective kernel and convolutional additive self-attention mamba for hyperspectral image classification |
| title_full_unstemmed | HLSK-CASMamba: hybrid large selective kernel and convolutional additive self-attention mamba for hyperspectral image classification |
| title_short | HLSK-CASMamba: hybrid large selective kernel and convolutional additive self-attention mamba for hyperspectral image classification |
| title_sort | hlsk casmamba hybrid large selective kernel and convolutional additive self attention mamba for hyperspectral image classification |
| topic | Hyperspectral image Classification Convolutional neural networks (CNNs) Large selective kernel Conv additive self-attention Mamba |
| url | https://doi.org/10.1007/s44443-025-00060-z |
| work_keys_str_mv | AT xiaoqingwan hlskcasmambahybridlargeselectivekernelandconvolutionaladditiveselfattentionmambaforhyperspectralimageclassification AT yupenghe hlskcasmambahybridlargeselectivekernelandconvolutionaladditiveselfattentionmambaforhyperspectralimageclassification AT fengchen hlskcasmambahybridlargeselectivekernelandconvolutionaladditiveselfattentionmambaforhyperspectralimageclassification AT ziqisun hlskcasmambahybridlargeselectivekernelandconvolutionaladditiveselfattentionmambaforhyperspectralimageclassification AT dongtaomo hlskcasmambahybridlargeselectivekernelandconvolutionaladditiveselfattentionmambaforhyperspectralimageclassification |