Visible to the public Generative Adversarial Network with Folded Spectrum for Hyperspectral Image Classification

TitleGenerative Adversarial Network with Folded Spectrum for Hyperspectral Image Classification
Publication TypeConference Paper
Year of Publication2019
AuthorsLi, Wenyue, Yin, Jihao, Han, Bingnan, Zhu, Hongmei
Conference NameIGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium
Date Publishedjul
Keywords2D square spectrum, abundant spectral information, classification accuracy, Classification algorithms, fake folded spectrum, feature extraction, feature matching strategy, folded spectrum, FS-GAN, Generative Adversarial Learning, generative adversarial network, generative adversarial networks, Generators, Hyperspectral classification, hyperspectral image classification, Hyperspectral imaging, image classification, labeled dataset, learning (artificial intelligence), Metrics, neural nets, nonadjacent spectral bands, original spectral vector, pubcrawl, resilience, Resiliency, Scalability, semi-supervised learning, Semisupervised learning, Shape, spectral texture, spectral-based classification methods

Hyperspectral image (HSIs) with abundant spectral information but limited labeled dataset endows the rationality and necessity of semi-supervised spectral-based classification methods. Where, the utilizing approach of spectral information is significant to classification accuracy. In this paper, we propose a novel semi-supervised method based on generative adversarial network (GAN) with folded spectrum (FS-GAN). Specifically, the original spectral vector is folded to 2D square spectrum as input of GAN, which can generate spectral texture and provide larger receptive field over both adjacent and non-adjacent spectral bands for deep feature extraction. The generated fake folded spectrum, the labeled and unlabeled real folded spectrum are then fed to the discriminator for semi-supervised learning. A feature matching strategy is applied to prevent model collapse. Extensive experimental comparisons demonstrate the effectiveness of the proposed method.

Citation Keyli_generative_2019