Two-Branch Convolutional Neural Network with Polarized Full Attention for Hyperspectral Image Classification
Open Access
- 2 February 2023
- journal article
- research article
- Published by MDPI AG in Remote Sensing
- Vol. 15 (3) , 848
Abstract
In recent years, convolutional neural networks (CNNs) have been introduced for pixel-wise hyperspectral image (HSI) classification tasks. However, some problems of the CNNs are still insufficiently addressed, such as the receptive field problem, small sample problem, and feature fusion problem. To tackle the above problems, we proposed a two-branch convolutional neural network with a polarized full attention mechanism for HSI classification. In the proposed network, two-branch CNNs are implemented to efficiently extract the spectral and spatial features, respectively. The kernel sizes of the convolutional layers are simplified to reduce the complexity of the network. This approach can make the network easier to be trained and fit the network to small sample size conditions. The one-shot connection technique is applied to improve the efficiency of feature extraction. An improved full attention block, named polarized full attention, is exploited to fuse the feature maps and provide global contextual information. Experimental results on several public HSI datasets confirm the effectiveness of the proposed network.Keywords
Funding Information
- National Natural Science Foundation of China (62071084, 145109218)
- Leading Talents Project of the State Ethnic Affairs Commission (62071084, 145109218)
- the Fundamental Research Funds in Heilongjiang Provincial Universities (62071084, 145109218)
This publication has 71 references indexed in Scilit:
- Hyperspectral Image Classification Using Deep Pixel-Pair FeaturesIEEE Transactions on Geoscience and Remote Sensing, 2016
- Spectral–Spatial Feature Extraction for Hyperspectral Image Classification: A Dimension Reduction and Deep Learning ApproachIEEE Transactions on Geoscience and Remote Sensing, 2016
- Generating labeled samples for hyperspectral image classification using correlation of spectral bandsFrontiers of Computer Science, 2015
- Hyperspectral and LiDAR Data Fusion: Outcome of the 2013 GRSS Data Fusion ContestIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014
- ${{\rm E}^{2}}{\rm LMs}$: Ensemble Extreme Learning Machines for Hyperspectral Image ClassificationIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014
- Hyperspectral Image Classification With Independent Component Discriminant AnalysisIEEE Transactions on Geoscience and Remote Sensing, 2011
- Spectral–Spatial Hyperspectral Image Segmentation Using Subspace Multinomial Logistic Regression and Markov Random FieldsIEEE Transactions on Geoscience and Remote Sensing, 2011
- Classification of Hyperspectral Images With Regularized Linear Discriminant AnalysisIEEE Transactions on Geoscience and Remote Sensing, 2009
- Classification of hyperspectral remote sensing images with support vector machinesIEEE Transactions on Geoscience and Remote Sensing, 2004
- Digital selection and analogue amplification coexist in a cortex-inspired silicon circuitNature, 2000