CNN‐combined graph residual network with multilevel feature fusion for hyperspectral image classification

Although graph convolutional network (GCN) has achieved remarkable success in hyperspectral image (HSI) classification, most existing GCN-based approaches have failed to realize a deep network structure due to the oversmoothing problem. This problem largely limits the expression ability and feature extraction ability of GCN and hampers GCN's capacity to model long-range relationships between samples in hyperspectral (HS) scenes. Moreover, there is a lack of theoretical analysis in those works that constructed deep GCN for HSI classification to illustrate how they overcome the oversmoothing problem. Aside from this, the characteristics and complexity of HSI are often neglected when constructing deep GCN models in HSI classification. To address these problems, a novel deep graph network based on first-order smoothing is proposed for HSI classification. Specifically, a local and global topologically consistent graph is constructed to thoroughly explore the union between fine pixel information and semantic superpixel information. Subsequently, a novel propagation procedure is proposed to address the oversmoothing problem. We creatively build a residual connection to the first layer to emphasize the feature information aggregated from the first-order neighborhood, which adds node features that have not yet become indistinguishable into deep layer, and at the same time, it can be considered as a correction to the original pixels affected by spectral variation in the input graph. Finally, we demonstrate how first-order smoothing-based deep graph network (FSDGN) can slow down the convergence rate of the oversmoothing problem by analyzing the propagation of FSDGN from the standpoint of the Laplacian spectral domain. In addition, the results of experiments performed on three benchmark datasets demonstrate its superiority over other state-of-the-art methods.

Recently, deep learning-based methods have made great progress in hyperspectral image (HSI) classification (HSIC). Different from ordinary images, the intrinsic complexity of HSIs data still limits the performance of many common convolutional neural network (CNN) models. Thus, the network architecture becomes more and more complex to extract discriminative spectral-spatial features. For instance, 3-D CNN usually has a large number of trainable parameters, thus increasing the computational complexity of the HSIC. In this letter, we designed a cooperative spectral-spatial attention dense network (CS <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ADN) that takes raw 3-D HSI data as input data. Specifically, the attention module consists of spectral and spatial axes, by which the salient spectral-spatial features will be emphasized. Furthermore, we combined these attention modules with the dense connection, which is termed as the lightweight dense block; it has a lower computation cost and achieves better classification performance. At the same time, we introduced the center loss, by jointly using the supervision of the center loss and the softmax loss, where the discriminative features could be clearly observed, particularly for small data sets. Experimental results on the biased and unbiased HSI data show that our method outperforms several state-of-the-art methods in HSIC with small training samples.

Adaptive spectral-spatial feature fusion network for hyperspectral image classification using limited training samples

Deep learning methods, especially convolutional neural networks(CNN), have been widely used in hyperspectral image(HSI) classification. Recently, graph convolutional networks (GCN) have shown great potential in HSI classification problem. However, the existing GCN-based methods have several problems. First, the existing methods rely too much on the adjacency matrix, which cannot be changed during training. Furthermore, most of them can only use a single kind of feature, and fail to extract the spectral-spatial information from the HSI. Finally, for the existing GCN-based methods, it is difficult to achieve the same accuracy as the mature CNN methods. In this paper, we propose a spectral-spatial hypergraph convolutional neural network (S²HCN) for HSI classification. Compared with the existing GCN-based methods, S²HCN has the following advantages. Different from the adjacency matrix that is fixed during training of GCN, S²HCN can dynamically update the weight of the hyperedge during training, which reduces the reliance on prior information to a certain extent. In addition, S²HCN generates hyperedges from the spectral and spatial features independently, and adopts the incidence matrix composed of all hyperedges to replace the adjacency matrix in GCN. In this way, the spectral and spatial features can be better integrated. Finally, compared to a simple graph structure, the hypergraph structure can express the high-dimensional relationships in the data, which is beneficial to classification problems. Sufficient experiments on two popular HSI datasets have proved the effectiveness of S²HCN.

Recently, hybrid networks, combining graph convolutional networks (GCNs) and convolutional neural networks (CNNs) into a unified framework, have garnered significant attention in hyperspectral image (HSI) classification. However, existing hybrid networks have the following limitations: 1) Existing methods primarily utilize simple fusion strategies such as concatenation or direct addition, resulting in the ineffective utilization of advantageous features. 2) Traditional GCNs only consider the relationship between pairs of vertices, limiting their ability to capture complex high-order and long-range correlations. In this work, a novel differential feature fusion network (DF2Net) is proposed for HSI classification. Specifically, DF2Net utilizes two subnetworks to learn features at different abstraction levels: the spectral-spatial hypergraph convolutional network (S2HGCN) for capturing complex high-order and long-range correlations, and the spectral-spatial convolution network (S2CN) for pixel-level local information extraction. Subsequently, we introduce an advantageous feature differential enhancement fusion (AFDEF) module, in which mutual enhancement of advantageous features from different network structures is performed, thereby improving the classification robustness of different regions in HSI. The experiments on four HSI benchmark datasets demonstrate that our DF2Net exhibits superior advantages over state-of-the-art models, particularly when the training samples are limited.

In recent years, deep-learning-based hyperspectral image (HSI) classification networks have become one of the most dominant implementations in HSI classification tasks. Among these networks, convolutional neural networks (CNNs) and attention-based networks have prevailed over other HSI classification networks. While convolutional neural networks with perceptual fields can effectively extract local features in the spatial dimension of HSI, they are poor at capturing the global and sequential features of spectral–spatial information; networks based on attention mechanisms, for example, Transformer, usually have better ability to capture global features, but are relatively weak in discriminating local features. This paper proposes a fusion network of convolution and Transformer for HSI classification, known as FusionNet, in which convolution and Transformer are fused in both serial and parallel mechanisms to achieve the full utilization of HSI features. Experimental results demonstrate that the proposed network has superior classification results compared to previous similar networks, and performs relatively well even on a small amount of training data.

Hyperspectral image (HSI) classification is one of the most crucial tasks in remote sensing processing. The attention mechanism is preferable to a convolutional neural network (CNN), due to its superior ability to express information during HSI processing. Recently, numerous methods combining CNNs and attention mechanisms have been applied in HSI classification. However, it remains a challenge to achieve high-accuracy classification by fully extracting effective features from HSIs under the conditions of limited labeled samples. In this paper, we design a novel HSI classification network based on multiscale hybrid networks and attention mechanisms. The network consists of three subnetworks: a spectral-spatial feature extraction network, a spatial inverted pyramid network, and a classification network, which are employed to extract spectral-spatial features, to extract spatial features, and to obtain classification results, respectively. The multiscale fusion network and attention mechanisms complement each other by capturing local and global features separately. In the spatial pyramid network, multiscale spaces are formed through down-sampling, which can reduce redundant information while retaining important information. The structure helps the network better capture spatial features at different scales, and to improve classification accuracy. Experimental results on various public HSI datasets demonstrate that the designed network is extremely competitive compared to current advanced approaches, under the condition of insufficient samples.

Fuzzy graph convolutional network for hyperspectral image classification

Hyperspectral image (HSI) classification has been extensively applied for analyzing remotely sensed images. HSI data consist of multiple bands that provide abundant spatial information. Convolutional neural networks (CNNs) have emerged as powerful deep learning methods for processing visual data. In recent work, CNNs have shown impressive results in HSI classification. In this paper, we propose a hierarchical neural network architecture called feature extraction with hybrid spectral CNN (FE-HybridSN) to extract superior spectral–spatial features. FE-HybridSN effectively captures more spectral–spatial information while reducing computational complexity. Considering the prevalent issue of class imbalance in experimental datasets (IP, UP, SV) and real-world hyperspectral datasets, we apply the focal loss to mitigate these problems. The focal loss reconstructs the loss function and facilitates effective achievement of the aforementioned goals. We propose a framework (FEHN-FL) that combines FE-HybridSN and the focal loss for HSI classification and then conduct extensive HSI classification experiments using three remote sensing datasets: Indian Pines (IP), University of Pavia (UP), and Salinas Scene (SV). Using cross-entropy loss as a baseline, we assess the hyperspectral classification performance of various backbone networks and examine the influence of different spatial sizes on classification accuracy. After incorporating focal loss as our loss function, we not only compare the classification performance of the FE-HybridSN backbone network under different loss functions but also evaluate their convergence rates during training. The proposed classification framework demonstrates satisfactory performance compared to state-of-the-art end-to-end deep-learning-based methods, such as 2D-CNN, 3D-CNN, etc.

Recently, graph convolutional network (GCN) has drawn increasing attention in hyperspectral image (HSI) classification, as it can process arbitrary non-Euclidean data. However, dynamic GCN that refines the graph heavily relies on the graph embedding in the previous layer, which will result in performance degradation when the embedding contains noise. In this letter, we propose a novel dual residual graph convolutional network (DRGCN) for HSI classification that integrates two adjacency matrices of dual GCN. In detail, one GCN applies a soft adjacency matrix to extract spatial features, the other utilizes the dynamic adjacency matrix to extract global context-aware features. Subsequently, the features extracted by dual GCN are fused to make full use of the complementary and correlated information among two graph representations. Moreover, we introduce residual learning to optimize graph convolutional layers during the training process, to alleviate the over-smoothing problem. The advantage of dual GCN is that it can extract robust and discriminative features from HSI. Extensive experiments on four HSI data sets, including Indian Pines, Pavia University, Salinas, and Houston University, demonstrate the effectiveness and superiority of our proposed DRGCN, even with small-sized training data.

Hyperspectral image (HSI) classification is fundamental to a wide range of remote sensing applications, such as precision agriculture, environmental monitoring, and urban planning, because HSIs provide rich spectral signatures that enable the discrimination of subtle material differences. Deep learning approaches, including Convolutional Neural Networks (CNNs), Graph Convolutional Networks (GCNs), and Transformers, have achieved strong performance in learning spatial–spectral representations. However, these models often face difficulties in jointly modeling long-range dependencies, fine-grained local structures, and non-Euclidean spatial relationships, particularly when labeled training data are scarce. This paper proposes a Spectral–Spatial Graph Transformer Network (SSGTN), a dual-branch architecture that integrates superpixel-based graph modeling with Transformer-based global reasoning. SSGTN consists of four key components, namely (1) an LDA-SLIC superpixel graph construction module that preserves discriminative spectral–spatial structures while reducing computational complexity, (2) a lightweight spectral denoising module based on 1×1 convolutions and batch normalization to suppress redundant and noisy bands, (3) a Spectral–Spatial Shift Module (SSSM) that enables efficient multi-scale feature fusion through channel-wise and spatial-wise shift operations, and (4) a dual-branch GCN-Transformer block that jointly models local graph topology and global spectral–spatial dependencies. Extensive experiments on three public HSI datasets (Indian Pines, WHU-Hi-LongKou, and Houston2018) under limited supervision (1% training samples) demonstrate that SSGTN consistently outperforms state-of-the-art CNN-, Transformer-, Mamba-, and GCN-based methods in overall accuracy, Average Accuracy, and the κ coefficient. The proposed framework provides an effective baseline for HSI classification under limited supervision and highlights the benefits of integrating graph-based structural priors with global contextual modeling.

Although most deep learning-based methods have achieved excellent performance for hyperspectral image (HSI) classification, they are often limited by complex networks and require massive training samples in practical applications. Therefore, designing an efficient, lightweight model to obtain better classification results under small samples situations remains a challenging task. To alleviate this problem, a novel, lightweight parallel octave convolutional neural network (LPOCNN) for HSI classification is proposed in this paper. First, the HSI data is preprocessed to construct two three-dimensional (3D) patch cubes with different spatial and spectral scales for each central pixel, removing redundancy and focusing on extracting spatial features and spectral features, respectively. Next, two non- deep parallel branches are created for the two inputs, which design octave convolution rather than classical 3D convolution to facilitate light weighting of the model. Then two-dimensional convolutional neural network is used to extract deeper spectral-spatial features when fusing spectral-spatial features from different parallel layers. Moreover, the spectral-spatial attention is designed to promote the classification performance even further by adaptively adjusting the weights of different spectral-spatial features according to their contribution to classification. Experiments show that our suggested LPOCNN acquires a significant advantage on classification performance over other competitive methods under small sample situations.

Hyperspectral image (HSI) classification has become one of the most significant tasks in the field of hyperspectral analysis. However, classifying each pixel in HSI accurately is challenging due to the curse of dimensionality and limited training samples. In this paper, we present an HSI classification architecture called camera spectral response network (CSR-Net), which can learn the optimal camera spectral response (CSR) function for HSI classification problems and effectively reduce the spectral dimensions of HSI. Specifically, we design a convolutional layer to simulate the capturing process of cameras, which learns the optimal CSR function for HSI classification. Then, spectral and spatial features are further extracted by spectral and spatial attention modules. On one hand, the learned CSR can be implemented physically and directly used to capture scenes, which makes the image acquisition process more convenient. On the other hand, compared with ordinary HSIs, we only need images with far fewer bands, without sacrificing the classification precision and avoiding the curse of dimensionality. The experimental results of four popular public hyperspectral datasets show that our method, with only a few image bands, outperforms state-of-the-art HSI classification methods which utilize the full spectral bands of images.

Convolutional neural networks (CNNs), a kind of feedforward neural network with a deep structure, are one of the representative methods in hyperspectral image (HSI) classification. However, redundant information and interclass interference are common and challenging problems in HSI classification. In addition, if the spectral and spatial information is not properly extracted and analyzed, it will affect the classification performance of the network to a great extent. Aiming at these issues, this article proposes an HSI classification method based on an adaptive hash attention mechanism and a lower triangular network (AHA-LT). First, the attention mechanism is introduced in the preprocessing stage, which is composed of the spectral attention module and the adaptive hash spatial attention module in series. Then, the data processed by the attention mechanism are introduced into the lower triangular network (LTNet) to obtain the fused high-dimensional semantic features. Finally, we compress the features and obtain the output classification results through several fully connected layers. Among them, LTNet is composed of 2-D–3-D CNN and multiscale features. The network integrates the characteristics of multibranch, feature fusion, feature compression, and skip connections. Extensive experiments on four widely used HSI data sets show that the proposed method can obtain a great improvement in performance compared with the existing methods.

The application of Transformer in computer vision has had the most significant influence of all the deep learning developments over the past five years. In addition to the exceptional performance of convolutional neural networks (CNN) in hyperspectral image (HSI) classification, Transformer has begun to be applied to HSI classification. However, for the time being, Transformer has not produced satisfactory results in HSI classification. Recently, in the field of image classification, the creators of Sequencer have proposed a Sequencer structure that substitutes the Transformer self-attention layer with a BiLSTM2D layer and achieves satisfactory results. As a result, this paper proposes a unique network called SquconvNet, that combines CNN with Sequencer block to improve hyperspectral classification. In this paper, we conducted rigorous HSI classification experiments on three relevant baseline datasets to evaluate the performance of the proposed method. The experimental results show that our proposed method has clear advantages in terms of classification accuracy and stability.