Generative Adversarial Networks-Based Semi-Supervised Learning for Hyperspectral Image Classification

Abstract

Classification of hyperspectral image (HSI) is an important research topic in the remote sensing community. Significant efforts (e.g., deep learning) have been concentrated on this task. However, it is still an open issue to classify the high-dimensional HSI with a limited number of training samples. In this paper, we propose a semi-supervised HSI classification method inspired by the generative adversarial networks (GANs). Unlike the supervised methods, the proposed HSI classification method is semi-supervised, which can make full use of the limited labeled samples as well as the sufficient unlabeled samples. Core ideas of the proposed method are twofold. First, the three-dimensional bilateral filter (3DBF) is adopted to extract the spectral-spatial features by naturally treating the HSI as a volumetric dataset. The spatial information is integrated into the extracted features by 3DBF, which is propitious to the subsequent classification step. Second, GANs are trained on the spectral-spatial features for semi-supervised learning. A GAN contains two neural networks (i.e., generator and discriminator) trained in opposition to one another. The semi-supervised learning is achieved by adding samples from the generator to the features and increasing the dimension of the classifier output. Experimental results obtained on three benchmark HSI datasets have confirmed the effectiveness of the proposed method, especially with a limited number of labeled samples.

Quality prediction via semisupervised Bayesian regression with application to propylene polymerization

“…Generative methods aim to estimate unlabeled data by a generative model that finds the label that maximizes the conditional probability with respect to the process states. Compared to other semisupervised method, the generative technique can provide suitable estimated values for unlabeled samples based on statistical properties of the industrial process …”

“…Compared to other semisupervised method, the generative technique can provide suitable estimated values for unlabeled samples based on statistical properties of the industrial process. [34][35][36] Hence, a semisupervised Bayesian prediction model is proposed in this paper. A Bayesian inference framework is developed to predict the quality variable, ie, MI for the polypropylene products, while neighborhood kernel density estimation is used to produce the relationship between the unlabeled data and labeled samples.…”

Siamese-GAN: Learning Invariant Representations for Aerial Vehicle Image Categorization

“…The model consists of a generator that learns to produce additional training images similar to the real data, and a discriminator that works as a feature extractor, which learns better representations of the images using the data provided by the generator. In another work, He et al [44] proposed a semi-supervised method for the classification of hyperspectral images. Spectral-spatial features are extracted from the unlabeled images and are used to train a GAN model.…”

“…For instance, Deep Convolutional GANs (DCGANs) [38] were designed to allow the network to generate data with similar internal structure as training data, improving the quality of the generated images, and Conditional GANs [39] add an additional conditioning variable to both the generator and the discriminator. Based on the previous architectures the concept of GANs has been adopted to solve many computer visions related tasks such as image generation [40,41], image super-resolution [42], unsupervised learning [43], semi-supervised learning [44], and image painting and colorization [45,46].…”

3D-Gabor Inspired Multiview Active Learning for Spectral-Spatial Hyperspectral Image Classification

“…However, among those methods, sufficient labeled samples/pixels are crucial to get the reliable classification results [18]. Since it is difficult to obtain a large number of labeled samples due to the time-consuming and expensive manual labeling process [19], defining a set of high informative training set is one of the solutions.…”