Learned Hyperspectral Compression Using a Student’s T Hyperprior

Hyperspectral compression is one of the most common techniques in hyperspectral image processing. Most recent learned image compression methods have exhibited excellent rate-distortion performance for natural images, but they have not been fully explored for hyperspectral compression tasks. In this...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Yuanyuan Guo, Yanwen Chong, Yun Ding, Shaoming Pan, Xiaolin Gu
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	artificial neural networks entropy model hyperspectral compression student’s T distribution Science Q
Acceso en línea:	https://doaj.org/article/bad49450caac49589fd4181c8fbd0e72
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:bad49450caac49589fd4181c8fbd0e72
record_format	dspace
spelling	oai:doaj.org-article:bad49450caac49589fd4181c8fbd0e722021-11-11T18:55:16ZLearned Hyperspectral Compression Using a Student’s T Hyperprior10.3390/rs132143902072-4292https://doaj.org/article/bad49450caac49589fd4181c8fbd0e722021-10-01T00:00:00Zhttps://www.mdpi.com/2072-4292/13/21/4390https://doaj.org/toc/2072-4292Hyperspectral compression is one of the most common techniques in hyperspectral image processing. Most recent learned image compression methods have exhibited excellent rate-distortion performance for natural images, but they have not been fully explored for hyperspectral compression tasks. In this paper, we propose a trainable network architecture for hyperspectral compression tasks, which not only considers the anisotropic characteristic of hyperspectral images but also embeds an accurate entropy model using the non-Gaussian prior knowledge of hyperspectral images and nonlinear transform. Specifically, we first design a spatial-spectral block, involving a spatial net and a spectral net as the base components of the core autoencoder, which is more consistent with the anisotropic hyperspectral cubes than the existing compression methods based on deep learning. Then, we design a Student’s T hyperprior that merges the statistics of the latents and the side information concepts into a unified neural network to provide an accurate entropy model used for entropy coding. This not only remarkably enhances the flexibility of the entropy model by adjusting various values of the degree of freedom, but also leads to a superior rate-distortion performance. The results illustrate that the proposed compression scheme supersedes the Gaussian hyperprior universally for virtually all learned natural image codecs and the optimal linear transform coding methods for hyperspectral compression. Specifically, the proposed method provides a 1.51% to 59.95% average increase in peak signal-to-noise ratio, a 0.17% to 18.17% average increase in the structural similarity index metric and a 6.15% to 64.60% average reduction in spectral angle mapping over three public hyperspectral datasets compared to the Gaussian hyperprior and the optimal linear transform coding methods.Yuanyuan GuoYanwen ChongYun DingShaoming PanXiaolin GuMDPI AGarticleartificial neural networksentropy modelhyperspectral compressionstudent’s T distributionScienceQENRemote Sensing, Vol 13, Iss 4390, p 4390 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	artificial neural networks entropy model hyperspectral compression student’s T distribution Science Q
spellingShingle	artificial neural networks entropy model hyperspectral compression student’s T distribution Science Q Yuanyuan Guo Yanwen Chong Yun Ding Shaoming Pan Xiaolin Gu Learned Hyperspectral Compression Using a Student’s T Hyperprior
description	Hyperspectral compression is one of the most common techniques in hyperspectral image processing. Most recent learned image compression methods have exhibited excellent rate-distortion performance for natural images, but they have not been fully explored for hyperspectral compression tasks. In this paper, we propose a trainable network architecture for hyperspectral compression tasks, which not only considers the anisotropic characteristic of hyperspectral images but also embeds an accurate entropy model using the non-Gaussian prior knowledge of hyperspectral images and nonlinear transform. Specifically, we first design a spatial-spectral block, involving a spatial net and a spectral net as the base components of the core autoencoder, which is more consistent with the anisotropic hyperspectral cubes than the existing compression methods based on deep learning. Then, we design a Student’s T hyperprior that merges the statistics of the latents and the side information concepts into a unified neural network to provide an accurate entropy model used for entropy coding. This not only remarkably enhances the flexibility of the entropy model by adjusting various values of the degree of freedom, but also leads to a superior rate-distortion performance. The results illustrate that the proposed compression scheme supersedes the Gaussian hyperprior universally for virtually all learned natural image codecs and the optimal linear transform coding methods for hyperspectral compression. Specifically, the proposed method provides a 1.51% to 59.95% average increase in peak signal-to-noise ratio, a 0.17% to 18.17% average increase in the structural similarity index metric and a 6.15% to 64.60% average reduction in spectral angle mapping over three public hyperspectral datasets compared to the Gaussian hyperprior and the optimal linear transform coding methods.
format	article
author	Yuanyuan Guo Yanwen Chong Yun Ding Shaoming Pan Xiaolin Gu
author_facet	Yuanyuan Guo Yanwen Chong Yun Ding Shaoming Pan Xiaolin Gu
author_sort	Yuanyuan Guo
title	Learned Hyperspectral Compression Using a Student’s T Hyperprior
title_short	Learned Hyperspectral Compression Using a Student’s T Hyperprior
title_full	Learned Hyperspectral Compression Using a Student’s T Hyperprior
title_fullStr	Learned Hyperspectral Compression Using a Student’s T Hyperprior
title_full_unstemmed	Learned Hyperspectral Compression Using a Student’s T Hyperprior
title_sort	learned hyperspectral compression using a student’s t hyperprior
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/bad49450caac49589fd4181c8fbd0e72
work_keys_str_mv	AT yuanyuanguo learnedhyperspectralcompressionusingastudentsthyperprior AT yanwenchong learnedhyperspectralcompressionusingastudentsthyperprior AT yunding learnedhyperspectralcompressionusingastudentsthyperprior AT shaomingpan learnedhyperspectralcompressionusingastudentsthyperprior AT xiaolingu learnedhyperspectralcompressionusingastudentsthyperprior
_version_	1718431668362018816

Learned Hyperspectral Compression Using a Student’s T Hyperprior

Ejemplares similares