Abstract:The band selection (BS) is an essential task in hyperspectral images (HSIs), consisting of huge spectral bands with noises and redundancies. The attention mechanism can be used for BS. However, the existing attention -based BS schemes are failed to capture the cross-dimension interaction between the input spectral and spatial dimensions during computation of attention weights and may produce poor feature representations. Again, the used reconstruction network in the existing BS methods unable to detect the HSIs features in multiple scales by blindly increasing the depth of the network. To deal with these problems, a novel end-to-end unsupervised triplet-attention multiscale reconstruction network for BS (TAttMSRecNet) has been proposed. The proposed network utilizes a triplet-attention mechanism having three parallel branches responsible to aggregate interactive cross-dimensional features between the spatial and spectral dimensions. After that, the network restores the original HSIs by using a 3D multiscale reconstruction network that applies multiple size convolution kernels to capture the discriminative HSIs features over the multiple scales where these features also communicate themselves to find the most efficacious HSIs information. In this way, the rich features are captured at a little computation cost, and the most informative bands can be effectively chosen for classification. Three standard data sets - Indian Pines (IP), Salinas (SA), and University of Pavia (UP) have been taken to conduct the experiments. The presented TAttMSRecNet can efficiently suppress the redundant or useless bands and selects more informative bands for better classification performance and also outperforms the other existing BS methods.

An Unsupervised Band Selection Method Based on Triple Constraints and Attention Network for Hyperspectral Imagery

A Band Selection Method with Masked Convolutional Autoencoder for Hyperspectral Image

An Unsupervised Band Selection Method Via Contrastive Learning for Hyperspectral Images

SSANet-BS: Spectral–Spatial Cross-Dimensional Attention Network for Hyperspectral Band Selection

Nonlocal Band Attention Network for Hyperspectral Image Band Selection

Unsupervised Band Selection for Hyperspectral Imagery Classification Without Manual Band Removal.

A Dense Spatial–spectral Attention Network for Hyperspectral Image Band Selection

TAttMSRecNet:Triplet-attention and multiscale reconstruction network for band selection in hyperspectral images

Unsupervised Hyperspectral Band Selection Based on Hypergraph Spectral Clustering

Hyperspectral Band Selection Using Pair-Wise Constraint And Band-Wise Correlation

An Unsupervised Band Selection Based on Band Similarity for Hyperspectral Image Target Detection.

BSFormer: Transformer-Based Reconstruction Network for Hyperspectral Band Selection

Target-Oriented Multi-criteria Band Selection for Hyperspectral Image.

BS-Nets: An End-to-End Framework For Band Selection of Hyperspectral Image

Attend in Bands: Hyperspectral Band Weighting and Selection for Image Classification

Unsupervised Band Selection Method Based on Importance-Assisted Column Subset Selection

Unsupervised band selection based on artificial bee colony algorithm for hyperspectral image classification.

EBARec-BS: Effective Band Attention Reconstruction Network for Hyperspectral Imagery Band Selection

Unsupervised Hyperspectral Image Band Selection Based on Deep Subspace Clustering.

Hyperspectral Band Selection with Convolutional Neural Network.

Unsupervised Band Selection Using Block-Diagonal Sparsity for Hyperspectral Image Classification.