TY - JOUR

T1 - SSA-SiamNet

T2 - Spectral-Spatial-Wise Attention-Based Siamese Network for Hyperspectral Image Change Detection

AU - Wang, L.

AU - Wang, Liguo

AU - Wang, Q.

AU - Atkinson, P.M.

N1 - ©2021 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

PY - 2022/1/31

Y1 - 2022/1/31

N2 - Deep learning methods, especially convolutional neural network (CNN)-based methods, have shown promising performance for hyperspectral image (HSI) change detection (CD). It is acknowledged widely that different spectral channels and spatial locations in input image patches may contribute differently to CD. However, they are treated equally in existing CNN-based approaches. To increase the accuracy of HSI CD, we propose an end-to-end Siamese CNN (SiamNet) with a spectral-spatial-wise attention (SSA-SiamNet) mechanism. The proposed SSA-SiamNet method can emphasize informative channels and locations and suppress less informative ones to refine the spectral-spatial features adaptively. Moreover, in the network training phase, the weighted contrastive loss function is used for more reliable separation of changed and unchanged pixels and to accelerate the convergence of the network. SSA-SiamNet was validated using four groups of bitemporal HSIs. The accuracy of CD using the SSA-SiamNet was found to be consistently greater than for ten benchmark methods.

AB - Deep learning methods, especially convolutional neural network (CNN)-based methods, have shown promising performance for hyperspectral image (HSI) change detection (CD). It is acknowledged widely that different spectral channels and spatial locations in input image patches may contribute differently to CD. However, they are treated equally in existing CNN-based approaches. To increase the accuracy of HSI CD, we propose an end-to-end Siamese CNN (SiamNet) with a spectral-spatial-wise attention (SSA-SiamNet) mechanism. The proposed SSA-SiamNet method can emphasize informative channels and locations and suppress less informative ones to refine the spectral-spatial features adaptively. Moreover, in the network training phase, the weighted contrastive loss function is used for more reliable separation of changed and unchanged pixels and to accelerate the convergence of the network. SSA-SiamNet was validated using four groups of bitemporal HSIs. The accuracy of CD using the SSA-SiamNet was found to be consistently greater than for ten benchmark methods.

KW - Attention mechanism

KW - change detection (CD)

KW - Convolution

KW - convolutional block attention module (CBAM)

KW - Feature extraction

KW - hyperspectral images (HSIs)

KW - Hyperspectral imaging

KW - Redundancy

KW - Siamese network (SiamNet).

KW - Task analysis

KW - Tensors

KW - Training

KW - Deep learning

KW - Learning systems

KW - Spectroscopy

KW - Change detection

KW - Image change detection

KW - Learning methods

KW - Network training

KW - Spatial features

KW - Spatial location

KW - Spectral channels

KW - Unchanged pixels

KW - Convolutional neural networks

U2 - 10.1109/TGRS.2021.3095899

DO - 10.1109/TGRS.2021.3095899

M3 - Journal article

VL - 60

JO - IEEE Transactions on Geoscience and Remote Sensing

JF - IEEE Transactions on Geoscience and Remote Sensing

SN - 0196-2892

M1 - 5510018

ER -