HyperGCN – a multi-layer multi-exit graph neural network to enhance hyperspectral image classification

With the continuous innovation in deep learning, it has become a major direction for scholars to introduce the knowledge of deep learning into hyperspectral image classification to enhance its classification accuracy. Convolutional Neural Networks (CNN) are one of the most commonly used deep learning-based visual data processing methods, and are widely used in hyperspectral image (HSI) classification by virtue of their excellent contextual modeling capability. Since the performance of HSI classification is highly dependent on spatial and spectral information, this paper proposes a hyperspectral image classification method using 3D attention mechanism in collaboration with Transformer for hyperspectral image classification in view of the problems that the current hyperspectral image classification models with the framework of CNN have insufficient spatial spectral feature extraction and fail to excavate and represent the sequence properties of spectral features well. In this paper, we introduce a variant Transformer model based on a hybrid model of both improved 3D-CNN and 2D-CNN, combining complementary information of spatial spectrum and spectra in the form of 3D convolution and 2D convolution on CNN, and adding a variant attention mechanism module to strengthen spatial texture features, while combining grouped transfer Transformer to jump connection to enable the lower layer to better learn the upper layer features. Firstly, a variant channel attention mechanism is introduced on 3D-CNN to enhance the acquisition of spectral information of image features by 3D-CNN. Secondly, a variant spatial attention mechanism is introduced to enable 3D-CNN to better acquire the spatial information of hyperspectral images in the network, and subsequently the acquired spatial and spectral feature information is passed to 2D-CNN to enable it to better acquire local feature information. Finally, the acquired image feature information is passed to the variant Transformer model to make up for the fact that CNN can only acquire hyperspectral image features in local contexts, enabling it to better acquire global feature information on feature sequences. The experimental results show that the proposed model is experimented on two hyperspectral datasets, Indian Pines and Pavia University, and the overall classification accuracy (OA), average classification accuracy (AA), and Kappa coefficient reach up to 99.59%, 99.31%, and 99.45%, respectively, on the PU dataset, compared with the current cutting-edge techniques. The classification accuracy has been improved.

In recent years, the use of deep neural network in effective network feature extraction and the design of efficient and high-precision hyperspectral image classification algorithms has gradually become a research hotspot for scholars. However, due to the difficulty of obtaining hyperspectral images and the high cost of annotation, the training samples are very limited. In order to cope with the small sample problem, researchers often deepen the network model and use the attention mechanism to extract features; however, as the network model continues to deepen, the gradient disappears, the feature extraction ability is insufficient, and the computational cost is high. Therefore, how to make full use of the spectral and spatial information in limited samples has gradually become a difficult problem. In order to cope with such problems, this paper proposes two-branch multiscale spatial–spectral feature aggregation with a self-attention mechanism for a hyperspectral image classification model (FHDANet); the model constructs a dense two-branch pyramid structure, which can achieve the high efficiency extraction of joint spatial–spectral feature information and spectral feature information, reduce feature loss to a large extent, and strengthen the model’s ability to extract contextual information. A channel–space attention module, ECBAM, is proposed, which greatly improves the extraction ability of the model for salient features, and a spatial information extraction module based on the deep feature fusion strategy HLDFF is proposed, which fully strengthens feature reusability and mitigates the feature loss problem brought about by the deepening of the model. Compared with five hyperspectral image classification algorithms, SVM, SSRN, A2S2K-ResNet, HyBridSN, SSDGL, RSSGL and LANet, this method significantly improves the classification performance on four representative datasets. Experiments have demonstrated that FHDANet can better extract and utilise the spatial and spectral information in hyperspectral images with excellent classification performance under small sample conditions.

Hyperspectral image (HSI) classification is one of the major problems in the field of remote sensing. Particularly, graph-based HSI classification is a promising topic and has received increasing attention in recent years. However, graphs with pixels as nodes generate large size graphs, thus increasing the computational burden. Moreover, satisfactory classification results are often not obtained without considering spatial information in constructing graph. To address these issues, this study proposes an efficient and effective semi-supervised spectral-spatial HSI classification method based on sparse superpixel graph (SSG). In the constructed sparse superpixels graph, each vertex represents a superpixel instead of a pixel, which greatly reduces the size of graph. Meanwhile, both spectral information and spatial structure are considered by using superpixel, local spatial connection and global spectral connection. To verify the effectiveness of the proposed method, three real hyperspectral images, Indian Pines, Pavia University and Salinas, are chosen to test the performance of our proposal. Experimental results show that the proposed method has good classification completion on the three benchmarks. Compared with several competitive superpixel-based HSI classification approaches, the method has the advantages of high classification accuracy (>97.85%) and rapid implementation (<10 s). This clearly favors the application of the proposed method in practice.

The rich spatial and spectral information in hyperspectral images (HSIs) makes spectral-spatial relationships essential for HSI classification (HSIC). Recent advancements indicate convolutional neural networks (CNNs) excel in HSIC but often struggle with precise spectral feature extraction. Moreover, the abundance of spectral information presents challenges in efficient feature representation and minimizing cross-domain interference. To address these limitations, we propose an efficient sequential spectral-spatial feature convolution network (S3FCN), employing successive subnetworks for spectral and spatial feature extraction with depthwise separable convolution. This approach balances the preservation of deep spectral and spatial features while significantly reducing network parameters, enhancing both performance and computational efficiency. We also introduce a sequential spectral-spatial attention module (S3AM) to integrate cross-domain correlations. This module utilizes spectral features from the preceding subnetwork and multilevel residual layers for in-depth exploration of spatial features, enabling deep integration for improved classification performance. The proposed architecture’s effectiveness is verified on five benchmark HSI datasets, including Pavia University, Salinas Valley, Kennedy Space Center, Indian Pines, and Houston 2013. Experimental results demonstrate that the sequential spectral-spatial connection in the feature extraction and attention mechanism integrated with depthwise separable convolution collectively surpasses current state-of-the-art (SOTA) techniques in classification accuracy with overall accuracies of 98.28%, 97.63%, 99.31%, 96.72%, and 95.38% across different datasets, while limiting the computation overhead, ensuring balanced network efficiency.

Convolutional Neural Networks (CNNs) have been extensively applied to hyperspectral (HS) image classification tasks and achieved promising performance. However, for CNN based HS image classification methods, it is hard to depict the dependencies among HS image pixels in long-range distanced positions and bands. Moreover, the limited receptive field of the convolutional layers extremely hinders the development of the CNN structure. To tackle these problems, in this paper, the novel Bottleneck Spatial-Spectral Transformer (BS2T) is proposed to depict the long-range global dependencies of HS image pixels, which can be regarded as a feature extraction module for HS image classification networks. More specifically, inspired by Bottleneck Transformer in computer vision, for HS image feature extraction, the proposed BS2T is incorporated with a feature contraction module, a multi-head spatial-spectral self-attention (MHS2A) module and a feature expansion module. In this way, convolutional operations are replaced by the MHS2A to capture the long-range dependency of HS pixels regardless of their spatial position and distance. Meanwhile, in the MHS2A module, to highlight the spectral features of HS images, we introduce the spectral information and content spatial positional information to classical multi-head self-attentions to make the attentions more positional aware and spectral aware. On this basis, a dual-branch HS image classification framework based on 3D CNN and BS2T is defined for jointly extracting the local-global features of HS images. Experimental results on three public HS image classification datasets show that the proposed classification framework achieves a significant improvement when comparing with the state-of-the-art methods. The source code of the proposed framework can be downloaded from https://github.com/srxlnnu/BS2T.

The adequate and finer spectral information in hyperspectral images (HSIs) are benefit for various downstream applications like smart agriculture and environmental monitoring. In HSI classification, dual-stream convolutional networks have gained much attention and have been widely used. In patch-based hyperspectral classification tasks, however, merely using center-labeled patches could lead to an increased unlabeled noise in the data. Moreover, in the application of dual-stream network structures, heterogeneity existed in both the data and feature semantic levels to capture more representative features. To tackle these challenges, we have devised a framework called cross-semantic heterogeneous modeling network (CreatingNet), which aligns more closely with the design principles of dual-stream networks by adjusting the input size. This framework introduces a distance metric attention mechanism (DMAM) based on spectral and spatial distances to strengthen the influence of the center pixel on the entire patch. Additionally, we present a fusion module named CrossViT, which combines features with diverse structures and characteristics, leveraging their complementarity. The proposed multiscale heterogeneous fusion module allows for more effective integration of spatial and spectral features in the images. Extensive experiments on four well-known HSI datasets (Indian Pines, Pavia University, Salinas, and Houston 2013) demonstrate the superior classification performance of the proposed CreatingNet to several state-of-the-art methods. The effectiveness of the proposed model is further validated through ablation studies.

In hyperspectral image (HSI) classification, convolutional neural networks (CNNs) have been widely employed and achieved promising performance. However, CNN-based methods face difficulties in achieving both accurate and efficient HSI classification due to their limited receptive fields and deep architectures. To alleviate these limitations, we propose an effective HSI classification network based on multi-head self-attention and spectral-coordinate attention (MSSCA). Specifically, we first reduce the redundant spectral information of HSI by using a point-wise convolution network (PCN) to enhance discriminability and robustness of the network. Then, we capture long-range dependencies among HSI pixels by introducing a modified multi-head self-attention (M-MHSA) model, which applies a down-sampling operation to alleviate the computing burden caused by the dot-product operation of MHSA. Furthermore, to enhance the performance of the proposed method, we introduce a lightweight spectral-coordinate attention fusion module. This module combines spectral attention (SA) and coordinate attention (CA) to enable the network to better weight the importance of useful bands and more accurately localize target objects. Importantly, our method achieves these improvements without increasing the complexity or computational cost of the network. To demonstrate the effectiveness of our proposed method, experiments were conducted on three classic HSI datasets: Indian Pines (IP), Pavia University (PU), and Salinas. The results show that our proposed method is highly competitive in terms of both efficiency and accuracy when compared to existing methods.

Hyperspectral images include richer spectral and spatial information than common images, which are widely used in military, agricultural fields, etc. With the development of sensor technology, the spatial resolution and spectral resolution of hyperspectral images have been improved significantly. However, the disadvantage that there may contain only one part of one object which has different spectral information in hyperspectral images. This will lead to unsatisfactory performance in traditional pixel-level hyperspectral image classification. Thus, a new hyperspectral image classification framework based on convolutional neural network is proposed. First, band selection is adopted to obtain multiple sets of false color images for small sample hyperspectral data. Then, parallel CNNs are introduced to get the classification results of different band combinations. Finally, statistical analysis strategy is performed to obtain the final output result. Experiments show that the classification accuracy of this method is better than that of the previous algorithm on the same dataset.

Convolutional Neural Network (CNN) is widely used in Hyperspectral Images (HSIs) classification. However, the fine-grained spatial (FGS) details are discarded during a sequence of convolution and pooling operations for most of CNN-based HSIs classification methods. To address this issue, a unified encoder-decoder framework is proposed to integrate high-level semantics and FGS details for HSIs classification, denoted by FGSCNN. The encoder, including a series of convolution and pooling layers, captures the high-level semantic information with low resolution feature maps. The decoder fuses the high-level low-resolution semantic and the fine-grained high-resolution spatial information, namely, to get the FGS features with high-level semantics. The deconvolution layers and skip connection are used in the decoder to retain the FGS details, while, convolution layers are also used to combine the FGS features with high-level semantics. Based on the encoder-decoder framework, a unified loss function is exploited to integrate the high-level semantic information and FGS details with an end-to-end manner for HSIs classification. Experiments conducted on the three public datasets, i.e. the Indian Pines, Pavia University and Salinas, demonstrate the effectiveness of the proposed method on HSIs classification.

Mamba-in-Mamba: Centralized Mamba-Cross-Scan in Tokenized Mamba Model for Hyperspectral image classification

Hyperspectral image (HSI) classification aims to distinguish the category of a land coverage object for each pixel. In an effective way, the transformer architecture has been successfully introduced for the HSI classification task with promising performance. However, existing transformer-based HSI classification methods still suffer from the inability to fully explore both spectral information and spatial information in HSIs. To this end, we propose a Spectral-Spatial Token Enhanced Transformer (SSTE-Former) method with the hash-based positional embedding, which is the first to exploit multiscale spectral-spatial information for transformer-based HSI classification in-depth. Specifically, SSTE-Former accepts multiscale HSI cubes centered on the target pixel, that are preprocessed by PCA. Then, a designed multiscale CNN architecture is utilized to extract short-range spectral-spatial features and generate token embeddings. In parallel, a novel hash-based spatially enhanced positional embedding tailored for HSI cubes is developed to model the correlations within and across multiscale token embeddings. Finally, multiscale token embeddings and hash-based positional embeddings are concatenated and flattened into the transformer encoder for long-range spectral-spatial feature fusion. We conduct extensive experiments on four benchmark HSI datasets and achieve superior performance compared with the state-of-the-art HSI classification methods.

Convolutional neural networks (CNNs) have attracted extensive attention in the field of modern remote sensing image processing and show outstanding performance in hyperspectral image (HSI) classification. Nevertheless, some hyperspectral images have fixed position priors and parameter sharing between different positions, so the common convolution layer may ignore some important fine and useful information and cannot guarantee to effectively capture the optimal image features. This paper proposes an improved multilayer perceptron (IMLP) and IMLP combined with ResNet (IMLP-ResNet) two models for HSI classification. Combined with the characteristics of hyperspectral data, we design IMLP based on three improvements. Specifically, a depthwise over-parameterized convolutional layer is introduced to increase learnable parameters of the model in IMLP, which speeds up the convergence of the model without increasing the computational complexity. Secondly, a Focal Loss function is used to suppress the useless ones in the classification task and enhance the critical spectral–spatial features, which allow the IMLP network to learn more useful hyperspectral image information. Furthermore, to enhance the convergence speed of the network, cosine annealing is introduced to further improve the training performance of IMLP. Furthermore, the IMLP module is combined with a residual network (IMLP-ResNet) to construct a symmetric structure, which extracts more advanced semantic information from hyperspectral images. The proposed IMLP and IMLP-ResNet are tested on the two public HSI datasets (i.e., Indian Pines and Pavia University) and a real hyperspectral dataset (Xuzhou). Experimental results demonstrate the superiority of the proposed IMLP-ResNet method over several state-of-the-art methods with the highest OA, which outperforms CNN by 8.19%, 6.28%, 5.59% and outperforms ResNet by 3.52%, 3.54%, 2.67% on Indian Pines, Pavia University and Xuzhou datasets, respectively, and demonstrates that the well-designed MLPs can also obtain remarkable classification performance of HSI.

The classification of hyperspectral images plays a critical role in the maintenance of remote image analysis, which has attracted a lot of research interest. Despite the fact that numerous methodologies, including unsupervised and supervised methods, have been presented, achieving an acceptable classification result remains a challenging task. Deep learning-based hyperspectral image (HSI) classification is gaining popularity, because of its efficient classification capabilities. When compared to traditional convolutional neural networks, graph-based deep learning provides the benefits of exhibiting class boundaries and modelling feature relationships. In hyperspectral image (HSI) classification, the most important problem is how to transform hyperspectral data into irregular domains from regular grids. This study describes a method for image classification that employs graph neural network (GNN) models. The input images are converted into region adjacency graphs (RAGs), where regions are super pixels and edges link nearby super pixels.

Creating accurate land use and land cover maps using remote sensing images is one of the most important applications of remotely sensed data. Abundant spectral information in hyperspectral images (HSI) makes it possible to distinguish materials that would not be distinguishable by multi-spectral sensors. Spectral and spatial information from HSI is of primary importance for image classification. In this study, a hybrid stacked autoencoder (SAE) architecture and support vector machine (SVM) classifier was constructed to classify the HSI. The SAE architecture is constituted by stacking a multiple autoencoder (AE) deep learning network that consists in the encoder and decoder process. Spatial features in a neighbor region extracted from the principal component analysis (PCA) and the texture feature extracted from the gray-level cooccurrence matrix (GLCM) were fed into the classifier. It was found that the best result was from the combination of GLCM texture feature, PCA spatial feature, and spectral feature. Meanwhile, the representative features derived from SAE deep learning network were better than the original features. It reminded us that extracting the representative features from hyperspectral images is a key step of improving classification accuracy.

Recently, graph convolutional network (GCN) has been applied for hyperspectral image (HSI) classification and obtained better performance. The main issue in HSI classification is that the high-resolution HSI contains more complex spectral-spatial structure information. However, the previous GCN-based methods applied in HSI classification only adopted a shallow GCN layer and they can not extract the deeper discriminative features. In addition, these methods ignored the complementary and correlated information among multi-order neighboring information extracted by multiple GCN layers. In this letter, a novel feature fusion via deep residual graph convolutional network is proposed to explore the internal relationship among HSI data. On the one hand, benefiting from residual learning to alleviate the over-smoothing problem, we can construct deep GCN layers to excavate deeper abstract features of HSI. On the other hand, we fuse the outputs of different GCN layers, and thus, the local structural information within multi-order neighborhood nodes can be fully utilized. Extensive experiments on four real HSI data sets, including Indian Pines, Pavia University, Salinas, and Houston University, demonstrate the superiority of the proposed method compared with other state-of-the-art methods in various evaluation criteria.