Abstract:Hyperspectral image (HSI) classification suffers from two serious problems, one is the limited labeled pixels, and the other is the class imbalance problem. As a result, the number of labeled pixels in many categories is not sufficient to characterize the spectral-spatial information, and train a satisfying deep model. By making full use of the information of unlabeled pixels, semi-supervised methods can provide better classification performance in the case of limited labeled pixels. However, they do not take into account the imbalance in the HSI data. As a method of data enhancement, generative adversarial networks focus on the above two problems and have also been widely used for the task of the HSI classification. In this work, we propose a distance constraints-based generative adversarial networks (DGAN) method for HSI classification to address these two problems. The DGAN employs the convolution autoencoder (AE) to extract the latent features of the HSI samples, and considers the reconstructed samples from the AE as the real samples for the later classifier and discriminator. In addition, the DGAN uses two distance constraints to solve the problems of the few labeled samples and class imbalance, the one latent-data distance constraint enforcing the generator to generate HSI samples for each class (especially the minority class), another discriminator-score distance constraint guiding the generator to synthesize samples that resemble the real HSI samples. Finally, the generated samples are combined classwise with the reconstructed samples and the real HSI samples to learn the parameters of the classifier and discriminator. Experimental results show that our method achieves state-of-the-art performance in terms of overall accuracy (OA) when trained with only 0.5%-4% of data sets from Indian Pines, Pavia University, and Botswana. Specifically, our method demonstrates improvements of 5.48%, 8.79%, and 0.91% on these three datasets, respectively. It reveals the gre- t potential of the DGAN model in generating the HSI samples for each class, which contributes to improving the classification performance of the HSI data.

Cascade conditional generative adversarial nets for spatial-spectral hyperspectral sample generation

SpatialGAN: Progressive Image Generation Based on Spatial Recursive Adversarial Expansion

CBS-GAN: A Band Selection Based Generative Adversarial Net for Hyperspectral Sample Generation

Sample Generation with Self-Attention Generative Adversarial Adaptation Network (SaGAAN) for Hyperspectral Image Classification

Spectral Image Visualization Using Generative Adversarial Networks

Generative Adversarial Networks and Probabilistic Graph Models for Hyperspectral Image Classification

Generative Adversarial Capsule Network With ConvLSTM for Hyperspectral Image Classification

Generative adversarial networks for data-scarce spectral applications

CVA2E: A Conditional Variational Autoencoder with an Adversarial Training Process for Hyperspectral Imagery Classification

Generating images for Supervised Hyperspectral Image Classification with Generative Adversarial Nets

Self-Supervised Divide-and-Conquer Generative Adversarial Network for Classification of Hyperspectral Images

Data augmentation using continuous conditional generative adversarial networks for regression and its application to improved spectral sensing

Cross-Scene Hyperspectral Image Classification Based on Cycle-Consistent Adversarial Networks.

Hyperspectral Image Classification Based on Generative Adversarial Networks with Feature Fusing and Dynamic Neighborhood Voting Mechanism

ECGAN: An Improved Conditional Generative Adversarial Network With Edge Detection to Augment Limited Training Data for the Classification of Remote Sensing Images With High Spatial Resolution

A Spatial Weather Generator Based on Conditional Deep Convolution Generative Adversarial Nets (Cdcgan)

Cross-Domain Hyperspectral Image Classification Based on Generative Adversarial Networks

Distance Constraint-Based Generative Adversarial Networks for Hyperspectral Image Classification

Two-Stage Domain Alignment Single-Source Domain Generalization Network for Cross-Scene Hyperspectral Images Classification

Semisupervised hyperspectral image classification based on generative adversarial networks and spectral angle distance

Sample generation based on a supervised Wasserstein Generative Adversarial Network for high-resolution remote-sensing scene classification