ASDN: A Deep Convolutional Network for Arbitrary Scale Image Super-Resolution

Shen, J; Wang, Y; Zhang, J

ASDN: A Deep Convolutional Network for Arbitrary Scale Image Super-Resolution

Shen, J Wang, Y Zhang, J

Permalink

Publisher:: Springer Science and Business Media LLC
Publication Type:: Journal Article
Citation:: Mobile Networks and Applications, 2021, 26, (1), pp. 13-26
Issue Date:: 2021-02-01

Closed Access

	Filename	Description	Size
	Shen2021_Article_ASDNADeepConvolutionalNetworkF.pdf	Published Version	3.71 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Shen, J
dc.contributor.author	Wang, Y
dc.contributor.author	Zhang, J https://orcid.org/0000-0002-7240-3541
dc.date.accessioned	2021-04-27T05:29:38Z
dc.date.available	2021-04-27T05:29:38Z
dc.date.issued	2021-02-01
dc.identifier.citation	Mobile Networks and Applications, 2021, 26, (1), pp. 13-26
dc.identifier.issn	1383-469X
dc.identifier.issn	1572-8153
dc.identifier.uri	http://hdl.handle.net/10453/148418
dc.description.abstract	Deep convolutional neural networks have significantly improved the peak signal-to-noise ratio of Super-Resolution (SR). However, image viewer applications commonly allow users to zoom the images to arbitrary magnification scales, thus far imposing a large number of required training scales at a tremendous computational cost. To obtain a more computationally efficient model for arbitrary-scale SR, this paper employs a Laplacian pyramid method to reconstruct any-scale high-resolution (HR) images using the high-frequency image details in a Laplacian Frequency Representation. For SR of small-scales (between 1 and 2), images are constructed by interpolation from a sparse set of precalculated Laplacian pyramid levels. SR of larger scales is computed by recursion from small scales, which significantly reduces the computational cost. For a full comparison, fixed- and any-scale experiments are conducted using various benchmarks. At fixed scales, ASDN outperforms predefined upsampling methods (e.g., SRCNN, VDSR, DRRN) by about 1 dB in PSNR. At any-scale, ASDN generally exceeds Meta-SR on many scales.
dc.language	en
dc.publisher	Springer Science and Business Media LLC
dc.relation.ispartof	Mobile Networks and Applications
dc.relation.isbasedon	10.1007/s11036-020-01720-2
dc.rights	info:eu-repo/semantics/closedAccess
dc.subject	0805 Distributed Computing, 0999 Other Engineering
dc.subject.classification	Networking & Telecommunications
dc.title	ASDN: A Deep Convolutional Network for Arbitrary Scale Image Super-Resolution
dc.type	Journal Article
utslib.citation.volume	26
utslib.for	0805 Distributed Computing
utslib.for	0999 Other Engineering
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Strength - GBDTC - Global Big Data Technologies
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology/School of Electrical and Data Engineering
utslib.copyright.status	closed_access	*
dc.date.updated	2021-04-27T05:29:34Z
pubs.issue	1
pubs.publication-status	Published
pubs.volume	26
utslib.citation.issue	1

Abstract:

Deep convolutional neural networks have significantly improved the peak signal-to-noise ratio of Super-Resolution (SR). However, image viewer applications commonly allow users to zoom the images to arbitrary magnification scales, thus far imposing a large number of required training scales at a tremendous computational cost. To obtain a more computationally efficient model for arbitrary-scale SR, this paper employs a Laplacian pyramid method to reconstruct any-scale high-resolution (HR) images using the high-frequency image details in a Laplacian Frequency Representation. For SR of small-scales (between 1 and 2), images are constructed by interpolation from a sparse set of precalculated Laplacian pyramid levels. SR of larger scales is computed by recursion from small scales, which significantly reduces the computational cost. For a full comparison, fixed- and any-scale experiments are conducted using various benchmarks. At fixed scales, ASDN outperforms predefined upsampling methods (e.g., SRCNN, VDSR, DRRN) by about 1 dB in PSNR. At any-scale, ASDN generally exceeds Meta-SR on many scales.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/148418