TY - JOUR
T1 - Eigenimage2Eigenimage (E2E)
T2 - A Self-Supervised Deep Learning Network for Hyperspectral Image Denoising
AU - Zhuang, Lina
AU - Ng, Michael K.
AU - Gao, Lianru
AU - Michalski, Joseph
AU - Wang, Zhicheng
N1 - This work was supported by the National Natural Science Foundation of China under Grant 42001287 and Grant 62161160336. The work of Michael K. Ng was supported in part by the Hong Kong Research Grants Council (HKRGC) General Research Fund (GRF) under Grant 12300519, Grant 17201020, and Grant 17300021; in part by the HKRGC Collaborative Research Fund (CRF) under Grant C1013-21GF and Grant C7004-21GF; and in part by the Joint NSFC and RGC under Grant N-HKU769/21. The work of Joseph Michalski was supported by the HKRGC CRF Fund under Grant C7004-21G.
Publisher Copyright:
IEEE
PY - 2024/11
Y1 - 2024/11
N2 - The performance of deep learning-based denoisers highly depends on the quantity and quality of training data. However, paired noisy–clean training images are generally unavailable in hyperspectral remote sensing areas. To solve this problem, this work resorts to the self-supervised learning technique, where our proposed model can train itself to learn one part of noisy input from another part of noisy input. We study a general hyperspectral image (HSI) denoising framework, called Eigenimage2Eigenimage (E2E), which turns the HSI denoising problem into an eigenimage (i.e., the subspace representation coefficients of the HSI) denoising problem and proposes a learning strategy to generate noisy–noisy paired training eigenimages from noisy eigenimages. Consequently, the E2E denoising framework can be trained without clean data and applied to denoise HSIs without the constraint with the number of frequency bands. Experimental results are provided to demonstrate the performance of the proposed method that is better than the other existing deep learning methods for denoising HSIs. A MATLAB demo of this work is available at https://github.com/LinaZhuang/HSI-denoiser-Eigenimage2Eigenimagehttps://github.com/LinaZhuang/HSI-denoiser-Eigenimage2Eigenimage for the sake of reproducibility.
AB - The performance of deep learning-based denoisers highly depends on the quantity and quality of training data. However, paired noisy–clean training images are generally unavailable in hyperspectral remote sensing areas. To solve this problem, this work resorts to the self-supervised learning technique, where our proposed model can train itself to learn one part of noisy input from another part of noisy input. We study a general hyperspectral image (HSI) denoising framework, called Eigenimage2Eigenimage (E2E), which turns the HSI denoising problem into an eigenimage (i.e., the subspace representation coefficients of the HSI) denoising problem and proposes a learning strategy to generate noisy–noisy paired training eigenimages from noisy eigenimages. Consequently, the E2E denoising framework can be trained without clean data and applied to denoise HSIs without the constraint with the number of frequency bands. Experimental results are provided to demonstrate the performance of the proposed method that is better than the other existing deep learning methods for denoising HSIs. A MATLAB demo of this work is available at https://github.com/LinaZhuang/HSI-denoiser-Eigenimage2Eigenimagehttps://github.com/LinaZhuang/HSI-denoiser-Eigenimage2Eigenimage for the sake of reproducibility.
KW - Convolutional neural network
KW - deep neural network
KW - hyperspectral image (HSI) restoration
KW - low-rank representation
KW - self-supervised learning
KW - subspace representation
UR - http://www.scopus.com/inward/record.url?scp=85165249970&partnerID=8YFLogxK
U2 - 10.1109/TNNLS.2023.3293328
DO - 10.1109/TNNLS.2023.3293328
M3 - Journal article
AN - SCOPUS:85165249970
SN - 2162-237X
VL - 35
SP - 16262
EP - 16276
JO - IEEE Transactions on Neural Networks and Learning Systems
JF - IEEE Transactions on Neural Networks and Learning Systems
IS - 11
ER -