Semi-supervised Hyperspectral Image Classification Research Articles

Dual-Stream Class-Adaptive Network for Semi-Supervised Hyperspectral Image Classification

Dual-Stream Class-Adaptive Network for Semi-Supervised Hyperspectral Image Classification

Dynamic evolution graph attention network for semi-supervised hyperspectral image classification

Dynamic evolution graph attention network for semi-supervised hyperspectral image classification

Adversarial MixUp with implicit semantic preservation for semi-supervised hyperspectral image classification

MixUp generates virtual samples by performing the identical convex combination in feature and semantic spaces simultaneously. However, convexly combining random pairs of images and their associated labels is likely to result in misunderstandings between the generated sample and its assigned pseudo label, particularly for hyperspectral images (HSI), leading to a confusing effect on the classifier. In this paper, adversarial MixUp with implicit semantic preservation (AdvMixup) is proposed for semi-supervised HSI classification (HSIc). The proposed AdvMixup exploits interpolation in the latent space for unlabeled samples. The virtual hyperspectral sample is generated by a deconvolutional network on the interpolated latent data. A semantic preservation network is performed on the newly generated virtual sample to recover the MixUp coefficient, where the semantic is preserved implicitly in the virtual sample. Meanwhile, an adversarial regularization is also conducted on the recovered mixup coefficient to ensure the authenticity of the virtual HSI. In such a way, the proposed AdvMixup sees the unseen HSI and knows the unknown semantic. Throughout the learning process, the feature extraction network is shared in both supervised and unsupervised branches to obtain a superior decision boundary. Experimental results on three HSI benchmark datasets demonstrate that AdvMixup achieves state-of-the-art performance in semi-supervised HSI classification.

Pseudolabel-Based Unreliable Sample Learning for Semi-Supervised Hyperspectral Image Classification

Pseudolabel-Based Unreliable Sample Learning for Semi-Supervised Hyperspectral Image Classification

Self-Supervised Assisted Semi-Supervised Residual Network for Hyperspectral Image Classification

Due to the scarcity and high cost of labeled hyperspectral image (HSI) samples, many deep learning methods driven by massive data cannot achieve the intended expectations. Semi-supervised and self-supervised algorithms have advantages in coping with this phenomenon. This paper primarily concentrates on applying self-supervised strategies to make strides in semi-supervised HSI classification. Notably, we design an effective and a unified self-supervised assisted semi-supervised residual network (SSRNet) framework for HSI classification. The SSRNet contains two branches, i.e., a semi-supervised and a self-supervised branch. The semi-supervised branch improves performance by introducing HSI data perturbation via a spectral feature shift. The self-supervised branch characterizes two auxiliary tasks, including masked bands reconstruction and spectral order forecast, to memorize the discriminative features of HSI. SSRNet can better explore unlabeled HSI samples and improve classification performance. Extensive experiments on four benchmarks datasets, including Indian Pines, Pavia University, Salinas, and Houston2013, yield an average overall classification accuracy of 81.65%, 89.38%, 93.47% and 83.93%, which sufficiently demonstrate that SSRNet can exceed expectations compared to state-of-the-art methods.

Semi-Supervised anchor graph ensemble for large-scale hyperspectral image classification

ABSTRACT Existing graph-based, semi-supervised hyperspectral image (HSI) classification models often suffer from prolonged execution time due to high computational complexity. In this work, we first propose a fast anchor graph regularization (FAGR) model for large scale, HSI classification. FAGR employs a simple anchor-based graph construction procedure and a new adjacency matrix among anchors to dramatically reduce the computational complexity while attaining good classification performance. In order to further improve the classification accuracy of hyperspectral images, we propose a novel semi-supervised anchor graph ensemble (SAGE) model. SAGE is an ensemble realization of multiple FAGR with each component FAGR operating on a randomly selected subset of features. Ameta classifier is applied to aggregate the outputs of component classifiers to yield an ensemble classification result. We performed extensive experimentations using three real-world HSI datasets, to compare the performance of FAGR and SAGE against several existing graph-based HSI classifiers. The experiment results show that the proposed SAGE achieves 95.78% classification accuracy on the Indian Pines dataset using limited labeled samples, out-performing existing models in terms of shorter execution time and better classification accuracy.

Dual Sparse Representation Graph-Based Copropagation for Semisupervised Hyperspectral Image Classification

Graph-based semisupervised hyperspectral image (HSI) classification methods have obtained extensive attention. In graph-based methods, a graph is first constructed, and then the label propagation is carried out on the constructed graph to obtain the labels for unknown samples. However, the results of label propagation may be unreliable, especially in the case of very limited labeled samples. To address the above problem, we propose dual sparse representation graph-based collaborative propagation (DSRG-CP) for HSI classification. Specifically, DSRG-CP adopts sparse representation (SR) to construct spectral and spatial graphs on spectral and spatial dimensions, respectively. Then, label propagation is performed on two graphs iteratively. In each iteration, only the samples with high classification confidence from one graph are added into another graph as labeled samples for the next label propagation. After several iterations, the labels of unlabeled samples are predicted by fusing the results of label propagation from two graphs. In addition, to make the spectral graph more discriminative, the regularizer of spectral statistical information is added into spectral SR model. To make the classification results more consistent in space, the superpixel block constraint is added into spatial graph model as regularizer. To evaluate the performance of the proposed method, DSRG-CP is compared with several graph-based methods and other state-of-the-art methods. Extensive experiments on real HSI data sets show that DSRG-CP can obtain competitive results for HSI classification.

Semi-Supervised classification of hyperspectral images using discrete nonlocal variation Potts Model

The classification of Hyperspectral Image (HSI) plays an important role in various fields. To achieve more precise multi-target classification in a short time, a method for combining discrete non-local theory with traditional variable fraction Potts models is presented in this paper. The nonlocal operator makes better use of the information in a certain region centered on that pixel. Meanwhile, adding the constraint in the model can ensure that every pixel in HSI has only one class. The proposed model has the characteristics of non-convex, nonlinear, and non-smooth so that it is difficult to achieve global optimization results. By introducing a series of auxiliary variables and using the alternating direction method of multipliers, the proposed classification model is transformed into a series of convex subproblems. Finally, we conducted comparison experiments with support vector machine (SVM), K-nearest neighbor (KNN), and convolutional neural network (CNN) on five different dimensional HSI data sets. The numerical results further illustrate that the proposed method is stable and efficient and our algorithm can get more accurate predictions in a shorter time, especially when classifying data sets with more spectral layers.

A Depthwise Separable Fully Convolutional ResNet With ConvCRF for Semisupervised Hyperspectral Image Classification

Hyperspectral images classification relies on the accurate and efficient extraction of discriminative features, detail preservation, and efficient learning with limited training samples. This article, therefore, presents an advanced neural network architecture combined with convolutional conditional random fields (ConvCRF) and region growing (RGW) approaches to address these key issues. First, a depthwise separable fully convolutional residual network (DFRes) is proposed for efficient feature learning, where a fully convolutional operation ensures a larger field of view, and residual learning and depthwise separable convolution can mitigate the problem of vanishing gradient and overfitting. Second, because the collection of ground-truth labels is usually difficult, the proposed architecture integrates the RGW method to effectively overcome the problem of limited training samples. Third, ConvCRF is used to preserve the image details for fine-grained predictions. Finally, the abovementioned key components are coherently integrated into the new semisupervised framework, i.e., DFRes with conditional random fields and RGW. Experimental results on three hyperspectral datasets demonstrate that the proposed approach outperforms other state-of-the-art methods.

Learning Dual Geometric Low-Rank Structure for Semisupervised Hyperspectral Image Classification.

Most of the available graph-based semisupervised hyperspectral image classification methods adopt the cluster assumption to construct a Laplacian regularizer. However, they sometimes fail due to the existence of mixed pixels whose recorded spectra are a combination of several materials. In this paper, we propose a geometric low-rank Laplacian regularized semisupervised classifier, by exploring both the global spectral geometric structure and local spatial geometric structure of hyperspectral data. A new geometric regularized Laplacian low-rank representation (GLapLRR)-based graph is developed to evaluate spectral-spatial affinity of mixed pixels. By revealing the global low-rank and local spatial structure of images via GLapLRR, the constructed graph has the characteristics of spatial-spectral geometry description, robustness, and low sparsity, from which a more accurate classification of mixed pixels can be achieved. The proposed method is experimentally evaluated on three real hyperspectral datasets, and the results show that the proposed method outperforms its counterparts, when only a small number of labeled instances are available.

Local Binary Ensemble based Self-training for Semi-supervised Classification of Hyperspectral Remote Sensing Images

Supervised classification of hyperspectral remote sensing images is still challenging due to the scarcity of enough labelled samples. Semi-supervised methods have been adopted to handle this issue. Self-training is a popular semi-supervised technique which is widely used for training a classifier with limited labelled data and a large quantity of unlabeled data. However, traditional self-training approaches often give poor classification results in high dimensional data. In the current work, a novel efficient self-training approach for handling the deficiency of labelled samples for semi-supervised classification of hyperspectral remote sensing images is proposed. The proposed method first trains an ensemble of locally specialized supervised binary classifiers independently by using the dimensionally reduced spectral feature vectors of few available labelled samples. The trained local binary classifiers are then used to extend the labelled set by iterative addition of highly informative unlabeled samples to it by exploiting both the spectral and spatial information of the hyperspectral image. The classifiers are then retrained with the extended dataset in a batchwise manner and the procedure is repeated until adequate quantity of labelled samples are generated. Finally, a supervised multiclass classifier is trained on the extended dataset to produce the final classification map. Experimental results on two benchmark hyperspectral image datasets prove the effectiveness of the proposed method over supervised and traditional self-training based semi-supervised pixelwise classification approach in terms of different classification measures.

Novel Semi-Supervised Hyperspectral Image Classification Based on a Superpixel Graph and Discrete Potential Method

Hyperspectral image (HSI) classification plays an important role in the automatic interpretation of the remotely sensed data. However, it is a non-trivial task to classify HSI accurately and rapidly due to its characteristics of having a large amount of data and massive noise points. To address this problem, in this work, a novel, semi-supervised, superpixel-level classification method for an HSI was proposed based on a graph and discrete potential (SSC-GDP). The key idea of the proposed scheme is the construction of the weighted connectivity graph and the division of the weighted graph. Based on the superpixel segmentation, a weighted connectivity graph is constructed usingthe weighted connection between a superpixel and its spatial neighbors. The generated graph is then divided into different communities/sub-graphs by using a discrete potential and the improved semi-supervised Wu–Huberman (ISWH) algorithm. Each community in the weighted connectivity graph represents a class in the HSI. The local connection strategy, together with the linear complexity of the ISWH algorithm, ensures the fast implementation of the suggested SSC-GDP method. To prove the effectiveness of the proposed spectral–spatial method, two public benchmarks, Indian Pines and Salinas, were utilized to test the performance of our proposal. The comparative test results confirmed that the proposed method was superior to several other state-of-the-art methods.

Semi-supervised classification of hyperspectral images based on two branch autoencoder

In this paper, we propose a novel semi-supervised method based on two branch autoencoder (TBAE) for hyperspectral images. A branch is classification function, and another branch is decoder function. Both of functions promote each other in training process. The encoder is fit for both classifier and decoder. The encoder and the decoder guide feature extraction from unlabeled samples. The encoder and the classifier guide classification for all labeled samples. We can get more information from unlabeled samples for classification. Finally, we use convolutional autoencoder (CAE) to extend TBAE and obtain TBCAE. Compared with ANN and CNN, TBAE and TBCAE have a better performance in the case of a small number of labeled samples. Use different datasets to verify these methods and print visualization of hidden layer and the reconstructed data. The results demonstrate that the proposed framework obtains credible results with a small number of labeled samples.

Semi-supervised hyperspectral image classification algorithm based on graph embedding and discriminative spatial information

Image classification is one of the important techniques in computer vision. Due to the limited access of labeled samples in hyperspectral images, semi-supervised learning (SSL) methods have been widely applied in hyperspectral image classification. Graph based semi-supervised learning provides an effective solution to model data in classification problems, of which graph construction is the critical step. In this paper we employ the graphs constructed with a typical manifold learning method-locally linear embedding (LLE), based on which semi-supervised classification is then conducted. To exploit the valuable spatial information contained in hyperspectral images, discriminative spatial information (DSI) is then extracted. The proposed classification method is evaluated using three real hyperspectral data sets, revealing state-of-art performance when compared with different classification methods.

Semi-Supervised Hyperspectral Image Classification via Spatial-Regulated Self-Training

Because there are many unlabeled samples in hyperspectral images and the cost of manual labeling is high, this paper adopts semi-supervised learning method to make full use of many unlabeled samples. In addition, those hyperspectral images contain much spectral information and the convolutional neural networks have great ability in representation learning. This paper proposes a novel semi-supervised hyperspectral image classification framework which utilizes self-training to gradually assign highly confident pseudo labels to unlabeled samples by clustering and employs spatial constraints to regulate self-training process. Spatial constraints are introduced to exploit the spatial consistency within the image to correct and re-assign the mistakenly classified pseudo labels. Through the process of self-training, the sample points of high confidence are gradually increase, and they are added to the corresponding semantic classes, which makes semantic constraints gradually enhanced. At the same time, the increase in high confidence pseudo labels also contributes to regional consistency within hyperspectral images, which highlights the role of spatial constraints and improves the HSIc efficiency. Extensive experiments in HSIc demonstrate the effectiveness, robustness, and high accuracy of our approach.

Semi-Supervised Hyperspectral Image Classification Based on Label Propagation via Selected Path

Most graph-based semi-supervised classification methods do not perform well in hyperspectral image classification tasks due to their high complexity and other limitations. This paper proposes a label propagation semi-supervised classification algorithm which uses a few selected important paths to considerably reduce computational costs. It establishes that the most important paths for label propagation are minimum cost paths. It proves that minimum cost paths exist in minimum cost trees (MCT), and proposes a method based on a variant minimum spanning tree (MST) combined with priority queue to construct MCTs. The algorithm propagates labels from unlabeled nodes to labeled ones, a unique way different from any other studies where propagation is in the opposite direction, which brings about several clear advantages. These include that only one propagation path is required for each unlabeled node, improving both timing and memory performance. It also helps to solve a problem posed by sparse graphs where some image pixels cannot be classified, a situation which is especially problematic in large-scale image classification. The proposed method has the advantages of linear computational complexity, is independent of data dimension, has fewer parameters and is insensitive to the values of parameters. Moreover, it does not need large numbers of labelled pixels nor complex training processes. Experiments on hyperspectral images have shown that, compared with several existing algorithms, the proposed method achieves better performance in less time. The paper addresses some fundamental issues regarding propagating labels in graph based semi-supervised classifications. Due to the simplicity and the fast speed of the algorithm, it is also suitable to be integrated into both state-of-the-art and future hyperspectral image classification frameworks which have a label propagation stage.

A Novel Tri-Training Technique for the Semi-Supervised Classification of Hyperspectral Images Based on Regularized Local Discriminant Embedding Feature Extraction

This paper introduces a novel semi-supervised tri-training classification algorithm based on regularized local discriminant embedding (RLDE) for hyperspectral imagery. In this algorithm, the RLDE method is used for optimal feature information extraction, to solve the problems of singular values and over-fitting, which are the main problems in the local discriminant embedding (LDE) and local Fisher discriminant analysis (LFDA) methods. An active learning method is then used to select the most useful and informative samples from the candidate set. In the experiments undertaken in this study, the three base classifiers were multinomial logistic regression (MLR), k-nearest neighbor (KNN), and random forest (RF). To confirm the effectiveness of the proposed RLDE method, experiments were conducted on two real hyperspectral datasets (Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) and Reflective Optics System Imaging Spectrometer (ROSIS)), and the proposed RLDE tri-training algorithm was compared with its counterparts of tri-training alone, LDE, and LFDA. The experiments confirmed that the proposed approach can effectively improve the classification accuracy for hyperspectral imagery.

Semi-Supervised hyperspectral image classification using local low-rank representation

ABSTRACTIn the area of hyperspectral image (HSI) classification, graph-based semi-supervised learning (SSL) has been proved to be highly effective. Constructing a proper graph is critical for graph-based SSL tasks. In HSI, spectral distance is widely used to calculate the weight of graph edge, though it can be influenced by noise and outliers. Meanwhile, links among all the data points are incorporated in the graph, including those from different subspaces. Thus the constructed graph might contain incorrect information. In this letter, a novel semi-supervised HSI classification method using local low-rank representation (SL2R) is proposed. Edge weight calculation will not be affected by noise or outliers thanks to the robustness of low-rank representation (LRR). Since each graph is constructed at local level, where pixels are basically embedded in the same subspace, links among uncorrelated pixels can be removed. Moreover, spatial context is naturally characterized by low-rank constraint on adjacent pixels. Experimental results on two data sets (Indian Pines and Botswana) confirm the effectiveness of the proposed method.

Spatial and class structure regularized sparse representation graph for semi-supervised hyperspectral image classification

Constructing a good graph that can capture intrinsic data structures is critical for graph-based semi-supervised learning methods, which are widely applied for hyperspectral image (HSI) classification with small amount of labeled samples.Among the existing graph construction methods, sparse representation (SR)-based methods have shown impressive performance on semi-supervised HSI classification tasks. However, most SR-based algorithms fail to consider the rich spatial information of HSI, which has been shown beneficial for classification tasks.In this paper, we propose a spatial and class structure regularized sparse representation (SCSSR) graph for semi-supervised HSI classification. Specifically, spatial information has been incorporated into SR model via the graph Laplacian regularization, it assumes that the spatial neighbors should have similar representation coefficients, the obtained coefficient matrix thus can reflect the similarity between samples more accurately. Besides, we also incorporate probabilistic class structure, which implies the probabilistic relationship between each sample and each class, into SR model to further improve discriminability of graph.The experimental results on Hyperion and AVIRIS hyperspectral data show that our method outperforms state of the art methods.

Semi-Supervised Classification of Hyperspectral Images Based on Extended Label Propagation and Rolling Guidance Filtering

Semi-supervised classification methods result in higher performance for hyperspectral images, because they can utilize the relationship between unlabeled samples and labeled samples to obtain pseudo-labeled samples. However, how generating an effective training sample set is a major challenge for semi-supervised methods, In this paper, we propose a novel semi-supervised classification method based on extended label propagation (ELP) and a rolling guidance filter (RGF) called ELP-RGF, in which ELP is a new two-step process to make full use of unlabeled samples. The first step is to implement the graph-based label propagation algorithm to propagate the label information from labeled samples to the neighboring unlabeled samples. This is then followed by the second step, which uses superpixel propagation to assign the same labels to all pixels within the superpixels that are generated by the image segmentation method, so that some labels wrongly labeled by the above step can be modified. As a result, so obtained pseudo-labeled samples could be used to improve the performance of the classifier. Subsequently, an effective feature extraction method, i.e., RGF is further used to remove the noise and the small texture structures to optimize the features of the initial hyperspectral image. Finally, these produced initial labeled samples and high-confidence pseudo-labeled samples are used as a training set for support vector machine (SVM). The experimental results show that the proposed method can produce better classification performance for three widely-used real hyperspectral datasets, particularly when the number of training samples is relatively small.