LSSMA: Lightweight Spectral–Spatial Neural Architecture With Multiattention Feature Extraction for Hyperspectral Image Classification

被引:2
|
作者
Ding, Shujie [1 ]
Ruan, Xiaoli [1 ]
Yang, Jing [1 ]
Sun, Jie [1 ]
Li, Shaobo [1 ]
Hu, Jianjun [2 ]
机构
[1] Guizhou Univ, State Key Lab Publ Big Data, Guiyang 550025, Peoples R China
[2] Univ South Carolina, Dept Comp Sci & Engn, Columbia, SC 29208 USA
关键词
Feature extraction; Hyperspectral imaging; Data mining; Image classification; Computational modeling; Data models; Context modeling; Hyperspectral image (HSI) classification; lightweight; multiattention feature extraction; spectral-spatial neural network; ATTENTION NETWORK; RESIDUAL NETWORK; GENERATION;
D O I
10.1109/JSTARS.2024.3371536
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Deep learning has been utilized for hyperspectral image (HSI) classification in recent years, with notable performance improvements. In particular, convolutional neural networks (CNNs) methods have achieved major advancements in this area. However, there are some drawbacks to the existing CNN-based HSI classification approaches: 1) the lack of effective and simple feature representations in CNNs, which overlook the effects of spectral differences and spatial contextual information; 2) the classification model has an enormous network complexity as a result of its numerous training parameters and high computational requirements; and 3) the category samples in HSI data exhibit a significant long tail distribution issue, which affects the classification performance of HSI. To address these issues, we propose a lightweight spectral-spatial neural architecture with multiattention feature extraction (LSSMA) for HSI classification. The main work consists of three areas: 1) A spectral feature extraction and fusing module is created to facilitate the fusing of spectral-spatial features while reducing the number of trainable parameters and computational complexity of the model. This module uses convolutions of various kernel sizes for residual connection and feature fusion, and introduces group convolution to achieve efficient feature representation. 2) To utilize the spectral-spatial correlation of HSI data to its fullest, a multiscale convolutional activation guided attention mechanism is designed and the position attention module is referenced, which can capture spectral differences and spatial contextual relationships between ground objects. and 3) Focal loss is applied in computer vision tasks to the LSSMA model to enhance its capacity to handle category imbalance. Experimental results on four publicly available hyperspectral datasets show that the method obtains better classification performance at a lower computational cost.
引用
收藏
页码:6394 / 6413
页数:20
相关论文
共 50 条
  • [21] Unseen Feature Extraction: Spatial Mapping Expansion With Spectral Compression Network for Hyperspectral Image Classification
    Yu, Chunyan
    Zhu, Yuanchen
    Song, Meiping
    Wang, Yulei
    Zhang, Qiang
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2024, 62
  • [22] A Lightweight Spectral-Spatial Convolution Module for Hyperspectral Image Classification
    Meng, Zhe
    Jiao, Licheng
    Liang, Miaomiao
    Zhao, Feng
    IEEE Geoscience and Remote Sensing Letters, 2022, 19
  • [23] Spatial Feature Enhancement and Attention-Guided Bidirectional Sequential Spectral Feature Extraction for Hyperspectral Image Classification
    Liu, Yi
    Jiang, Shanjiao
    Liu, Yijin
    Mu, Caihong
    REMOTE SENSING, 2024, 16 (17)
  • [24] Spectral Tensor Synthesis Analysis for Hyperspectral Image Spectral–Spatial Feature Extraction
    Ronghua Yan
    Jinye Peng
    Dongmei Ma
    Desheng Wen
    Journal of the Indian Society of Remote Sensing, 2019, 47 : 91 - 100
  • [25] SSFAN: A Compact and Efficient Spectral-Spatial Feature Extraction and Attention-Based Neural Network for Hyperspectral Image Classification
    Wang, Chunyang
    Zhan, Chao
    Lu, Bibo
    Yang, Wei
    Zhang, Yingjie
    Wang, Gaige
    Zhao, Zongze
    REMOTE SENSING, 2024, 16 (22)
  • [26] Feature Extraction for Hyperspectral Image Classification
    Uddin, M. P.
    Mamun, M. A.
    Hossain, M. A.
    2017 IEEE REGION 10 HUMANITARIAN TECHNOLOGY CONFERENCE (R10-HTC), 2017, : 379 - 382
  • [27] Multi-type spectral spatial feature for hyperspectral image classification
    Yuan, Yuan
    Jin, Mingxin
    NEUROCOMPUTING, 2022, 492 : 637 - 650
  • [28] Masked Spectral-Spatial Feature Prediction for Hyperspectral Image Classification
    Zhou, Feng
    Xu, Chao
    Yang, Guowei
    Hang, Renlong
    Liu, Qingshan
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2024, 62 : 1 - 13
  • [29] A SUBPIXEL SPATIAL-SPECTRAL FEATURE MINING FOR HYPERSPECTRAL IMAGE CLASSIFICATION
    Xu, Xiang
    Li, Jun
    Zhang, Yanning
    Li, Shutao
    IGARSS 2018 - 2018 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM, 2018, : 8476 - 8479
  • [30] Spectral-Spatial Discriminant Feature Learning for Hyperspectral Image Classification
    Dong, Chunhua
    Naghedolfeizi, Masoud
    Aberra, Dawit
    Zeng, Xiangyan
    REMOTE SENSING, 2019, 11 (13)