Packing Convolutional Neural Networks in the Frequency Domain

Wang, Y; Xu, C; Xu, C; Tao, D

Packing Convolutional Neural Networks in the Frequency Domain

Wang, Y Xu, C

Xu, C

Tao, D

Permalink

Publisher:: Institute of Electrical and Electronics Engineers
Publication Type:: Journal Article
Citation:: IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 41, (10), pp. 2495-2509
Issue Date:: 2019

Closed Access

	Filename	Description	Size
	08413170.pdf	Published version	1.22 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	Wang, Y
dc.contributor.author	Xu, C https://orcid.org/0000-0002-4756-0609
dc.contributor.author	Xu, C https://orcid.org/0000-0002-4756-0609
dc.contributor.author	Tao, D https://orcid.org/0000-0001-7225-5449
dc.date.accessioned	2020-05-07T06:13:47Z
dc.date.available	2020-05-07T06:13:47Z
dc.date.issued	2019
dc.identifier.citation	IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 41, (10), pp. 2495-2509
dc.identifier.issn	0162-8828
dc.identifier.issn	1939-3539
dc.identifier.uri	http://hdl.handle.net/10453/140543
dc.description.abstract	IEEE Deep convolutional neural networks (CNNs) are successfully used in a number of applications. However, their storage and computational requirements have largely prevented their widespread use on mobile devices. Here we present a series of approaches for compressing and speeding up CNNs in the frequency domain, which focuses not only on smaller weights but on all the weights and their underlying connections. By treating convolutional filters as images, we decompose their representations in the frequency domain as common parts (i.e., cluster centers) shared by other similar filters and their individual private parts (i.e., individual residuals). A large number of low-energy frequency coefficients in both parts can be discarded to produce high compression without significantly compromising accuracy. Furthermore, we explore a data-driven method for removing redundancies in both spatial and frequency domains, which allows us to discard more useless weights by keeping similar accuracies. After obtaining the optimal sparse CNN in the frequency domain, we relax the computational burden of convolution operations in CNNs by linearly combining the convolution responses of discrete cosine transform (DCT) bases. The compression and speed-up ratios of the proposed algorithm are thoroughly analyzed and evaluated on benchmark image datasets to demonstrate its superiority over state-of-the-art methods.
dc.format	Print-Electronic
dc.language	eng
dc.publisher	Institute of Electrical and Electronics Engineers
dc.relation.ispartof	IEEE Transactions on Pattern Analysis and Machine Intelligence
dc.relation.isbasedon	10.1109/TPAMI.2018.2857824
dc.rights	info:eu-repo/semantics/restrictedAccess
dc.subject	0801 Artificial Intelligence and Image Processing, 0806 Information Systems, 0906 Electrical and Electronic Engineering
dc.subject.classification	Artificial Intelligence & Image Processing
dc.title	Packing Convolutional Neural Networks in the Frequency Domain
dc.type	Journal Article
utslib.citation.volume	41
utslib.location.activity	United States
utslib.for	0801 Artificial Intelligence and Image Processing
utslib.for	0806 Information Systems
utslib.for	0906 Electrical and Electronic Engineering
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney
utslib.copyright.status	closed_access	*
pubs.consider-herdc	false
dc.date.updated	2020-05-07T06:13:40Z
pubs.issue	10
pubs.publication-status	Accepted
pubs.volume	41
utslib.start-page	2495
utslib.citation.issue	10

Abstract:

IEEE Deep convolutional neural networks (CNNs) are successfully used in a number of applications. However, their storage and computational requirements have largely prevented their widespread use on mobile devices. Here we present a series of approaches for compressing and speeding up CNNs in the frequency domain, which focuses not only on smaller weights but on all the weights and their underlying connections. By treating convolutional filters as images, we decompose their representations in the frequency domain as common parts (i.e., cluster centers) shared by other similar filters and their individual private parts (i.e., individual residuals). A large number of low-energy frequency coefficients in both parts can be discarded to produce high compression without significantly compromising accuracy. Furthermore, we explore a data-driven method for removing redundancies in both spatial and frequency domains, which allows us to discard more useless weights by keeping similar accuracies. After obtaining the optimal sparse CNN in the frequency domain, we relax the computational burden of convolution operations in CNNs by linearly combining the convolution responses of discrete cosine transform (DCT) bases. The compression and speed-up ratios of the proposed algorithm are thoroughly analyzed and evaluated on benchmark image datasets to demonstrate its superiority over state-of-the-art methods.

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

http://hdl.handle.net/10453/140543