Hyperspectral Image Denoising Using SURE-Based Unsupervised Convolutional Neural Networks

Han V. Nguyen; Magnus O. Ulfarsson; Johannes R. Sveinsson

doi:10.1109/TGRS.2020.3008844

Hyperspectral Image Denoising Using SURE-Based Unsupervised Convolutional Neural Networks

Han V. Nguyen, Magnus O. Ulfarsson, Johannes R. Sveinsson

Faculty of Electrical and Computer Engineering

University of Iceland

Research output: Contribution to journal › Article › peer-review

Abstract

Hyperspectral images (HSIs) are useful for many remote sensing applications. However, they are usually affected by noise that degrades the HSIs quality. Therefore, HSI denoising is important to improve the performance of subsequent HSI processing and analysis. In this article, we propose an HSI denoising method called Stein's unbiased risk estimate-convolutional neural network (SURE-CNN). The method is based on an unsupervised CNN and SURE. The main difference of SURE-CNN from existing supervised learning methods is that the SURE-based loss function can be computed only from noisy data. Since SURE is an unbiased estimate of the mean squared error (MSE) of an estimator, training a CNN using the SURE loss can yield similar results as using the MSE with ground truth in supervised learning. Also, a subspace version of SURE-CNN is proposed to reduce the running time. Extensive experimental results with both simulated and real data sets show that the SURE-CNN method outperforms the competitive methods in both objective and subjective assessments.

Original language	English
Article number	9146687
Pages (from-to)	3369-3382
Number of pages	14
Journal	IEEE Transactions on Geoscience and Remote Sensing
Volume	59
Issue number	4
DOIs	https://doi.org/10.1109/TGRS.2020.3008844
Publication status	Published - Apr 2021

Bibliographical note

Funding Information: Manuscript received March 24, 2020; revised June 1, 2020; accepted July 7, 2020. Date of publication July 23, 2020; date of current version March 25, 2021. This work was supported in part by the Icelandic Research Fund under Grant 174075-05 and Grant 207233-051, and in part by the University of Iceland Doctoral Fund under Grant 1547-154305. (Corresponding author: Magnus O. Ulfarsson.) The authors are with the Faculty of Electrical and Computer Engineering, University of Iceland, 107 Reykjavik, Iceland (e-mail: [email protected]; [email protected]). Publisher Copyright: © 1980-2012 IEEE.

Other keywords

Convolutional neural networks (CNNs)
Stein's unbiased risk estimate (SURE)
hyperspectral (HSI) image denoising
unsupervised deep learning (DL)

Access to Document

10.1109/TGRS.2020.3008844

Cite this

@article{1650be8f1436418e8a41a12f31eddd9d,

title = "Hyperspectral Image Denoising Using SURE-Based Unsupervised Convolutional Neural Networks",

abstract = "Hyperspectral images (HSIs) are useful for many remote sensing applications. However, they are usually affected by noise that degrades the HSIs quality. Therefore, HSI denoising is important to improve the performance of subsequent HSI processing and analysis. In this article, we propose an HSI denoising method called Stein's unbiased risk estimate-convolutional neural network (SURE-CNN). The method is based on an unsupervised CNN and SURE. The main difference of SURE-CNN from existing supervised learning methods is that the SURE-based loss function can be computed only from noisy data. Since SURE is an unbiased estimate of the mean squared error (MSE) of an estimator, training a CNN using the SURE loss can yield similar results as using the MSE with ground truth in supervised learning. Also, a subspace version of SURE-CNN is proposed to reduce the running time. Extensive experimental results with both simulated and real data sets show that the SURE-CNN method outperforms the competitive methods in both objective and subjective assessments. ",

keywords = "Convolutional neural networks (CNNs), Stein's unbiased risk estimate (SURE), hyperspectral (HSI) image denoising, unsupervised deep learning (DL)",

author = "Nguyen, \{Han V.\} and Ulfarsson, \{Magnus O.\} and Sveinsson, \{Johannes R.\}",

note = "Funding Information: Manuscript received March 24, 2020; revised June 1, 2020; accepted July 7, 2020. Date of publication July 23, 2020; date of current version March 25, 2021. This work was supported in part by the Icelandic Research Fund under Grant 174075-05 and Grant 207233-051, and in part by the University of Iceland Doctoral Fund under Grant 1547-154305. (Corresponding author: Magnus O. Ulfarsson.) The authors are with the Faculty of Electrical and Computer Engineering, University of Iceland, 107 Reykjavik, Iceland (e-mail: [email protected]; [email protected]). Publisher Copyright: {\textcopyright} 1980-2012 IEEE.",

year = "2021",

month = apr,

doi = "10.1109/TGRS.2020.3008844",

language = "English",

volume = "59",

pages = "3369--3382",

journal = "IEEE Transactions on Geoscience and Remote Sensing",

issn = "0196-2892",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "4",

}

TY - JOUR

T1 - Hyperspectral Image Denoising Using SURE-Based Unsupervised Convolutional Neural Networks

AU - Nguyen, Han V.

AU - Ulfarsson, Magnus O.

AU - Sveinsson, Johannes R.

N1 - Funding Information: Manuscript received March 24, 2020; revised June 1, 2020; accepted July 7, 2020. Date of publication July 23, 2020; date of current version March 25, 2021. This work was supported in part by the Icelandic Research Fund under Grant 174075-05 and Grant 207233-051, and in part by the University of Iceland Doctoral Fund under Grant 1547-154305. (Corresponding author: Magnus O. Ulfarsson.) The authors are with the Faculty of Electrical and Computer Engineering, University of Iceland, 107 Reykjavik, Iceland (e-mail: [email protected]; [email protected]). Publisher Copyright: © 1980-2012 IEEE.

PY - 2021/4

Y1 - 2021/4

N2 - Hyperspectral images (HSIs) are useful for many remote sensing applications. However, they are usually affected by noise that degrades the HSIs quality. Therefore, HSI denoising is important to improve the performance of subsequent HSI processing and analysis. In this article, we propose an HSI denoising method called Stein's unbiased risk estimate-convolutional neural network (SURE-CNN). The method is based on an unsupervised CNN and SURE. The main difference of SURE-CNN from existing supervised learning methods is that the SURE-based loss function can be computed only from noisy data. Since SURE is an unbiased estimate of the mean squared error (MSE) of an estimator, training a CNN using the SURE loss can yield similar results as using the MSE with ground truth in supervised learning. Also, a subspace version of SURE-CNN is proposed to reduce the running time. Extensive experimental results with both simulated and real data sets show that the SURE-CNN method outperforms the competitive methods in both objective and subjective assessments.

AB - Hyperspectral images (HSIs) are useful for many remote sensing applications. However, they are usually affected by noise that degrades the HSIs quality. Therefore, HSI denoising is important to improve the performance of subsequent HSI processing and analysis. In this article, we propose an HSI denoising method called Stein's unbiased risk estimate-convolutional neural network (SURE-CNN). The method is based on an unsupervised CNN and SURE. The main difference of SURE-CNN from existing supervised learning methods is that the SURE-based loss function can be computed only from noisy data. Since SURE is an unbiased estimate of the mean squared error (MSE) of an estimator, training a CNN using the SURE loss can yield similar results as using the MSE with ground truth in supervised learning. Also, a subspace version of SURE-CNN is proposed to reduce the running time. Extensive experimental results with both simulated and real data sets show that the SURE-CNN method outperforms the competitive methods in both objective and subjective assessments.

KW - Convolutional neural networks (CNNs)

KW - Stein's unbiased risk estimate (SURE)

KW - hyperspectral (HSI) image denoising

KW - unsupervised deep learning (DL)

UR - https://www.scopus.com/pages/publications/85103344643

U2 - 10.1109/TGRS.2020.3008844

DO - 10.1109/TGRS.2020.3008844

M3 - Article

SN - 0196-2892

VL - 59

SP - 3369

EP - 3382

JO - IEEE Transactions on Geoscience and Remote Sensing

JF - IEEE Transactions on Geoscience and Remote Sensing

IS - 4

M1 - 9146687

ER -

Hyperspectral Image Denoising Using SURE-Based Unsupervised Convolutional Neural Networks

Abstract

Bibliographical note

Other keywords

Access to Document

Fingerprint

Cite this