期刊
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
卷 19, 期 -, 页码 -出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/LGRS.2021.3127075
关键词
Hyperspectral imaging; Training; Deep learning; Loss measurement; Convolutional neural networks; Data models; Signal to noise ratio; Deep learning; hyperspectral unmixing; self-supervised learning
类别
资金
- Indian Institute of Information Technology Vadodara, India
This article introduces a two-stage fully connected self-supervised deep learning network for blind hyperspectral unmixing. The network jointly estimates endmembers and abundances, and learns the physics of hyperspectral image acquisition to alleviate practical issues. Experimental results show that the proposed model performs better than the state of the art in both qualitative and quantitative evaluations.
The deep learning methods have started showing promising results for spectral unmixing. We observe that many of them need direct supervision in the form of unmixed components, which is scarcely available in practice. At the same time, some require extensive data for supervised training in order to handle high implicit noise. In this letter, we propose a two-stage fully connected self-supervised deep learning network for alleviating these practical issues in performing blind hyperspectral unmixing. Given the data, the first stage (inverse model) jointly estimates the endmembers and abundances, whereas the second stage (forward model) learns the physics of hyperspectral image acquisition. The central idea is to reconstruct the hyperspectral input vector using estimated endmembers and abundances at the inverse model, which best presents the input vector's underlying physics to the forward model. The network is jointly optimized against a two-stage loss function (in measurement domain) during the training, and we decouple the second stage at the time of inference. AdamW is used to optimize the loss function, while ReLU with a dropout of 0.3 is employed to avoid possible overfitting. The proposed network requires only the sensed hyperspectral data and learns the unmixing function from data itself, even at a lower signal-to-noise ratio (SNR). Experiments are conducted on synthetic data at different SNRs and two real benchmark hyperspectral data. The efficacy of the proposed model is evaluated and compared with the state of the art both qualitatively and quantitatively.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据