Hyperspectral Image Classification Based on Generative Adversarial Networks with Feature Fusing and Dynamic Neighborhood Voting Mechanism
| Autor: | Xianchuan Yu, Yuntao Wang, Yasmine Medjadba, Ying Cao, Jin Qin, Kang Wu, Zhengang Zhao, Guian Wang, Dan Hu, Ying Zhan, RunCheng Jiao, Tao Huang |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Discriminator
Computer science business.industry media_common.quotation_subject Deep learning 0211 other engineering and technologies Hyperspectral imaging Pattern recognition 02 engineering and technology Adversarial system ComputingMethodologies_PATTERNRECOGNITION Voting 0202 electrical engineering electronic engineering information engineering Hyperspectral image classification 020201 artificial intelligence & image processing Artificial intelligence business Classifier (UML) Generative grammar 021101 geological & geomatics engineering media_common |
| Zdroj: | IGARSS |
| DOI: | 10.1109/igarss.2019.8899291 |
| Popis: | Classifying Hyperspectral images with few training samples is a challenging problem. The generative adversarial networks (GAN) are promising techniques to address the problems. GAN constructs an adversarial game between a discriminator and a generator. The generator generates samples that are not distinguishable by the discriminator, and the discriminator determines whether or not a sample is composed of real data. In this paper, by introducing multilayer features fusion in GAN and a dynamic neighborhood voting mechanism, a novel algorithm for HSIs classification based on 1-D GAN was proposed. Extracting and fusing multiple layers features in discriminator, and using a little labeled samples, we fine-tuned a new sample 1-D CNN spectral classifier for HSIs. In order to improve the accuracy of the classification, we proposed a dynamic neighborhood voting mechanism to classify the HSIs with spatial features. The obtained results show that the proposed models provide competitive results compared to the state-of-the-art methods. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|