Abstract:Supervised classification of hyperspectral image (HSI) is generally required to obtain better performance in spectral-spatial feature learning by fully using complex pixel- and superpixel-level interdependencies with small labeled samples. Limited by the local regular convolutions, convolutional neural networks (CNNs) can only exploit information from the short-range Euclidean neighbors of a target, hindering the effectiveness of feature representation. In contrast, graph convolutional networks (GCNs) can learn long-range dependencies between non-Euclidean neighbors but usually require the input of a full graph constructed from a whole HSI, making GCNs must be trained in a full-batch manner with tremendous computational consumption. In this work, we propose a composite neighbor-aware convolutional metric network (CNCMN), aiming to learn each target's representation from its composite neighbors (i.e., both Euclidean and non-Euclidean neighbors) in a batchwise manner. Specifically, for each target in an HSI, its Euclidean neighbors are the pixels in the local square region centered on itself, and its non-Euclidean neighbors are several related nodes selected from the constructed full graph. Correspondingly, a composite convolution (CoConv) is proposed by coupling an image convolution and a graph convolution, which can perform flexible convolutions on those composite neighbors and extract adaptively fused features from them. Besides, to further boost classification, we also propose a mini-batch metric classifier to dynamically optimize interclass and intraclass distances of samples batch by batch, which is then combined with the CoConv to form the mini-batch CNCMN. Extensive experiments on three real-world HSIs demonstrate the advantages of the proposed method over mini-batch deep learning algorithms and have obtained the state-of-the-art performance in these fields. The code is available at: https://github.com/qichaoliu/HSI-CNCMN.

Statistical Texture Awareness Network for Hyperspectral Image Classification

Learning Statistical Texture for Semantic Segmentation

Composite Neighbor-Aware Convolutional Metric Networks for Hyperspectral Image Classification

Multiscale Residual Weakly Dense Network with Attention Mechanism for Hyperspectral Image Classification

End-to-End Convolutional Network and Spectral-Spatial Transformer Architecture for Hyperspectral Image Classification

Structure-Texture Aware Network for Low-Light Image Enhancement

Multiscale Densely Connected Attention Network for Hyperspectral Image Classification

Deep Texture Recognition Via Exploiting Cross-Layer Statistical Self-Similarity

Hyperspectral Image Classification Based on Multilevel Joint Feature Extraction Network

Hyperspectral Image Classification with a Dense Multi-Scale Residual Network

HSI-CNN: A Novel Convolution Neural Network for Hyperspectral Image

Hyperspectral image classification based on multiscale piecewise spectral-spatial attention network

Convolution Transformer Fusion Splicing Network for Hyperspectral Image Classification

Structure- and Texture-Aware Learning for Low-Light Image Enhancement

Pseudo Complex-Valued Deformable ConvLSTM Neural Network With Mutual Attention Learning for Hyperspectral Image Classification

Hybrid Fully Connected Tensorized Compression Network for Hyperspectral Image Classification.

Hyperspectral image classification based on a novel Lush multi-layer feature fusion bias network

Hyperspectral Image Classification Based on a 3D Octave Convolution and 3D Multiscale Spatial Attention Network

Spatial Proximity Feature Selection With Residual Spatial-Spectral Attention Network for Hyperspectral Image Classification.

HDTFF-Net: Hierarchical Deep Texture Features Fusion Network for High-resolution Remote Sensing Scene Classification

An Effective Hyperspectral Image Classification Network Based on Multi-Head Self-Attention and Spectral-Coordinate Attention