Active Learning-Driven Siamese Network for Hyperspectral Image Classification

Hyperspectral image (HSI) classification is one of the main research contents of hyperspectral technology. Existing HSI classification algorithms that are based on deep learning use a large number of labeled samples to train models to ensure excellent classification effects, but when the labeled samples are insufficient, the deep learning model is prone to overfitting. In practice, there are a large number of unlabeled samples that have not been effectively utilized, so it is meaningful to study a semi-supervised method. In this paper, an adversarial representation learning that is based on a generative adversarial networks (ARL-GAN) method is proposed to solve the small samples problem in hyperspectral image classification by applying GAN to the representation learning domain in a semi-supervised manner. The proposed method has the following distinctive advantages. First, we build a hyperspectral image block generator whose input is the feature vector that is extracted from the encoder and use the encoder as a feature extractor to extract more discriminant information. Second, the distance of the class probability output by the discriminator is used to measure the error between the generated image block and the real image instead of the root mean square error (MSE), so that the encoder can extract more useful information for classification. Third, GAN and conditional entropy are used to improve the utilization of unlabeled data and solve the small sample problem in hyperspectral image classification. Experiments on three public datasets show that the method achieved better classification accuracy with a small number of labeled samples compared to other state-of-the-art methods.

Deep learning classifiers exhibit remarkable performance for hyperspectral image classification given sufficient labeled samples but show deficiency in the situation of learning with limited labeled samples. Active learning endows deep learning classifiers with the ability to alleviate this deficiency. However, existing active deep learning methods tend to underestimate the feature variability of hyperspectral images when querying informative unlabeled samples subject to certain acquisition heuristics. A major reason for this bias is that the acquisition heuristics are normally derived based on the output of a deep learning classifier, in which representational power is bounded by the number of labeled training samples at hand. To address this limitation, we developed a feature-oriented adversarial active learning (FAAL) strategy, which exploits the high-level features from one intermediate layer of a deep learning classifier for establishing an acquisition heuristic based on a generative adversarial network (GAN). Specifically, we developed a feature generator for generating fake high-level features and a feature discriminator for discriminating between the real high-level features and the fake ones. Trained with both the real and the fake high-level features, the feature discriminator comprehensively captures the feature variability of hyperspectral images and yields a powerful and generalized discriminative capability. We leverage the well-trained feature discriminator as the acquisition heuristic to measure the informativeness of unlabeled samples. Experimental results validate the effectiveness of both (i) the full FAAL framework and (ii) the adversarially learned acquisition heuristic, for the task of classifying hyperspectral images with limited labeled samples.

In recent years, deep neural networks have been widely used for hyperspectral image (HSI) classification and have shown excellent performance using numerous labeled samples. The acquisition of HSI labels is usually based on the field investigation, which is expensive and time consuming. Hence, the available labels are usually limited, which affects the efficiency of deep HSI classification methods. To improve the classification performance while reducing the labeling cost, this article proposes a semisupervised deep learning (DL) method for HSI classification, named pyramidal dilation attention convolutional network with active and self-paced learning (PDAC-ASPL), which integrates active learning (AL), self-paced learning (SPL), and DL into a unified framework. First, a densely connected pyramidal dilation attention convolutional network is trained with a limited number of labeled samples. Then, the most informative samples from the unlabeled set are selected by AL and queried real labels, and the highest confidence samples with corresponding pseudo labels are extracted by SPL. Finally, the samples from AL and SPL are added to the training set to retrain the network. Compared with some DL- and AL-based HSI classification methods, our PDAC-ASPL achieves better performance on four HSI datasets.

Hyperspectral image (HSI) classification has been one of the most important tasks in the remote sensing community over the last few decades. Due to the presence of highly correlated bands and limited training samples in HSI, discriminative feature extraction was challenging for traditional machine learning methods. Recently, deep learning based methods have been recognized as powerful feature extraction tool and have drawn a significant amount of attention in HSI classification. Among various deep learning models, convolutional neural networks (CNNs) have shown huge success and offered great potential to yield high performance in HSI classification. Motivated by this successful performance, this paper presents a systematic review of different CNN architectures for HSI classification and provides some future guidelines. To accomplish this, our study has taken a few important steps. First, we have focused on different CNN architectures, which are able to extract spectral, spatial, and joint spectral-spatial features. Then, many publications related to CNN based HSI classifications have been reviewed systematically. Further, a detailed comparative performance analysis has been presented between four CNN models namely 1D CNN, 2D CNN, 3D CNN, and feature fusion based CNN (FFCNN). Four benchmark HSI datasets have been used in our experiment for evaluating the performance. Finally, we concluded the paper with challenges on CNN based HSI classification and future guidelines that may help the researchers to work on HSI classification using CNN.

Due to a lack of labeled samples, deep learning methods generally tend to have poor classification performance in practical applications. Few-shot learning (FSL), as an emerging learning paradigm, has been widely utilized in hyperspectral image (HSI) classification with limited labeled samples. However, the existing FSL methods generally ignore the domain shift problem in cross-domain scenes and rarely explore the associations between samples in the source and target domain. To tackle the above issues, a graph-based domain adaptation FSL (GDAFSL) method is proposed for HSI classification with limited training samples, which utilizes the graph method to guide the domain adaptation learning process in a uniformed framework. First, a novel deep residual hybrid attention network (DRHAN) is designed to extract discriminative embedded features efficiently for few-shot HSI classification. Then, a graph-based domain adaptation network (GDAN), which combines graph construction with domain adversarial strategy, is proposed to fully explore the domain correlation between source and target embedded features. By utilizing the fully explored domain correlations to guide the domain adaptation process, a domain invariant feature metric space is learned for few-shot HSI classification. Comprehensive experimental results conducted on three public HSI datasets demonstrate that GDAFSL is superior to the state-of-the-art with a small sample size.

Active deep learning (ADL) presents an appropriate solution for hyperspectral images (HSIs) classification based on domain adaptation (DA) with limited labelled samples in the target domain. But some challenges still exist. First, traditional ADL methods only match the feature distributions between the source and target domains globally without considering the decision boundaries between classes, which makes the ambiguous features near land-cover class boundaries and reduces the classification accuracy. Second, in previous ADL settings, a trained classifier is first used to obtain the predictions for the unlabelled data and then a measure is applied to achieve the uncertainty of such a classifier prediction. This two-step approach does not consider unlabelled data in the classifier training, which ignores dealing with noisy and complex data in the target domain. To overcome these issues, we propose adversarial discriminative active deep learning (ADADL), which presents an adversarial model and incorporates two different land-cover classifiers as a discriminator to consider the class boundaries in aligning feature distributions. Furthermore, ADADL combines the entropy measure along with the cross-entropy loss during training to use the information on the unlabelled data of the target domain. The experimental results with two benchmark HSIs show that the proposed ADADL creates robust transferable features far from the original class boundaries and improves the classification accuracy significantly compared to the state-of-art ADL methods even in complex and noisy data.

The deep learning methods have recently been successfully explored for hyperspectral image classification. However, it may not perform well when training samples are scarce. A deep transfer learning method is proposed to improve the hyperspectral image classification performance in the situation of limited training samples. First, a Siamese network composed of two convolutional neural networks is designed for local image descriptors extraction. Subsequently, the pretrained Siamese network model is reused to transfer knowledge to the hyperspectral image classification tasks by feeding deep features extracted from each band into a recurrent neural network. Indeed, a deep convolutional recurrent neural network is constructed for hyperspectral image classification by this way. Finally, the entire network is tuned by a small number of labeled samples. The important characteristic of the designed model is that the deep convolutional recurrent neural network provides a way of utilizing the spatial–spectral features without dimension reduction. Furthermore, the transfer learning method provides an opportunity to train such deep model with limited labeled samples. Experiments on three widely used hyperspectral datasets demonstrate that the proposed transfer learning method can improve the classification performance and competitive classification results can be achieved when compared with state-of-the-art methods.

With the rapid development of deep learning technology and improvement in computing capability, deep learning has been widely used in the field of hyperspectral image (HSI) classification. In general, deep learning models often contain many trainable parameters and require a massive number of labeled samples to achieve optimal performance. However, in regard to HSI classification, a large number of labeled samples is generally difficult to acquire due to the difficulty and time-consuming nature of manual labeling. Therefore, many research works focus on building a deep learning model for HSI classification with few labeled samples. In this article, we concentrate on this topic and provide a systematic review of the relevant literature. Specifically, the contributions of this paper are twofold. First, the research progress of related methods is categorized according to the learning paradigm, including transfer learning, active learning and few-shot learning. Second, a number of experiments with various state-of-the-art approaches has been carried out, and the results are summarized to reveal the potential research directions. More importantly, it is notable that although there is a vast gap between deep learning models (that usually need sufficient labeled samples) and the HSI scenario with few labeled samples, the issues of small-sample sets can be well characterized by fusion of deep learning methods and related techniques, such as transfer learning and a lightweight model. For reproducibility, the source codes of the methods assessed in the paper can be found at https://github.com/ShuGuoJ/HSI-Classification.git.

Deep learning has recently attracted significant attention in the field of hyperspectral images (HSIs) classification. However, the construction of an efficient deep neural network (DNN) mostly relies on a large number of labeled samples being available. To address this problem, this paper proposes a unified deep network, combined with active transfer learning that can be well-trained for HSIs classification using only minimally labeled training data. More specifically, deep joint spectral-spatial feature is first extracted through hierarchical stacked sparse autoencoder (SSAE) networks. Active transfer learning is then exploited to transfer the pre-trained SSAE network and the limited training samples from the source domain to the target domain, where the SSAE network is subsequently fine-tuned using the limited labeled samples selected from both source and target domain by corresponding active learning strategies. The advantages of our proposed method are threefold: 1) the network can be effectively trained using only limited labeled samples with the help of novel active learning strategies; 2) the network is flexible and scalable enough to function across various transfer situations, including cross-dataset and intra-image; 3) the learned deep joint spectral-spatial feature representation is more generic and robust than many joint spectral-spatial feature representation. Extensive comparative evaluations demonstrate that our proposed method significantly outperforms many state-of-the-art approaches, including both traditional and deep network-based methods, on three popular datasets.

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, hyperspectral image (HSI) classification approaches based on deep learning (DL) models have been proposed and shown promising performance. However, because of very limited available training samples and massive model parameters, DL methods may suffer from overfitting. In this paper, we propose an end-to-end 3-D lightweight convolutional neural network (CNN) (abbreviated as 3-D-LWNet) for limited samples-based HSI classification. Compared with conventional 3-D-CNN models, the proposed 3-D-LWNet has a deeper network structure, less parameters, and lower computation cost, resulting in better classification performance. To further alleviate the small sample problem, we also propose two transfer learning strategies: 1) cross-sensor strategy, in which we pretrain a 3-D model in the source HSI data sets containing a greater number of labeled samples and then transfer it to the target HSI data sets and 2) cross-modal strategy, in which we pretrain a 3-D model in the 2-D RGB image data sets containing a large number of samples and then transfer it to the target HSI data sets. In contrast to previous approaches, we do not impose restrictions over the source data sets, in which they do not have to be collected by the same sensors as the target data sets. Experiments on three public HSI data sets captured by different sensors demonstrate that our model achieves competitive performance for HSI classification compared to several state-of-the-art methods

Multi-scale receptive fields: Graph attention neural network for hyperspectral image classification

Collaborative learning of lightweight convolutional neural network and deep clustering for hyperspectral image semi-supervised classification with limited training samples

Supervised hyperspectral image (HSI) classification has been widely studied and used in many different applications. However, the performance of the supervised classifiers, including the traditional machine learning methods and the deep neural networks, is significantly affected by the inaccurate labeling of training samples, which is a common problem in HSI supervised classification. In this article, we propose a superpixel guided sample selection neural network (S3Net) framework with end-to-end training for handling noisy labels in HSI classification. It includes two stages: sample selection and sample correction. In sample selection, a sample with a small training loss has a higher probability of being the correct label and hence selected from the noisy labels for model training. In order to avoid the error propagation caused by the noisy labels, we utilize a cross-selection update strategy that exchanges selected samples between two neural networks during conventional loss backpropagation. Sample selection is a pruning process, which may cause insufficient training sample problem in HSI classification. To solve this problem, we propose the sample correction strategy to correct the noisy labels by propagating clean label information in the homogeneous regions obtained by superpixel. Experimental results on three public HSI data sets demonstrate the effectiveness of the proposed S3Net framework when handling noisy labels.

In the field of hyperspectral image (HSI) classification, deep learning has helped achieve great successes. However, most of these achievements are made with very large amounts of labeled training data. Manual annotation of HSIs is labor intensive and time consuming. In practical HSI classification, there may only be a few labeled samples available. To perform HSI classification with a small number of labeled samples, a new few-shot classification model based on adaptive subspaces and featurewise transformation is proposed in this article. First, we design a 3-D local channel attention residual network to obtain the spatial–spectral features of HSIs. Then, a featurewise transformation strategy is introduced to enhance feature diversity to avoid model overfitting problems and to mitigate the impact of cross-domain problems. Finally, a subspace classifier is implemented to construct different subspace categories based on the embedded features of the limited labeled samples. Classification of an HSI sample is performed using spatial projection and a distance metric. The proposed model is trained using the metalearning mechanism to perform few-shot classification of HSIs. Four public datasets are utilized to construct a sufficient few-shot classification task named episodes for training. The other three public datasets are used to test the proposed model. Experiments show that our proposed method can outperform mainstream small sample HSI classification methods.