FCL-GAN: A Lightweight and Real-Time Baseline for Unsupervised Blind Image Deblurring

Zhao, S; Zhang, Z; Hong, R; Xu, M; Yang, Y; Wang, M

FCL-GAN: A Lightweight and Real-Time Baseline for Unsupervised Blind Image Deblurring

Zhao, S Zhang, Z Hong, R Xu, M Yang, Y

Wang, M

Permalink

Publisher:: Association for Computing Machinery (ACM)
Publication Type:: Conference Proceeding
Citation:: MM 2022 - Proceedings of the 30th ACM International Conference on Multimedia, 2022, pp. 6220-6229
Issue Date:: 2022-10-10

Closed Access

	Filename	Description	Size
	3503161.3548113_published.pdf	Published version	2.57 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	Zhao, S
dc.contributor.author	Zhang, Z
dc.contributor.author	Hong, R
dc.contributor.author	Xu, M
dc.contributor.author	Yang, Y https://orcid.org/0000-0002-0512-880X
dc.contributor.author	Wang, M
dc.date.accessioned	2023-04-17T05:57:58Z
dc.date.available	2023-04-17T05:57:58Z
dc.date.issued	2022-10-10
dc.identifier.citation	MM 2022 - Proceedings of the 30th ACM International Conference on Multimedia, 2022, pp. 6220-6229
dc.identifier.isbn	9781450392037
dc.identifier.uri	http://hdl.handle.net/10453/169860
dc.description.abstract	Blind image deblurring (BID) remains a challenging and significant task. Benefiting from the strong fitting ability of deep learning, paired data-driven supervised BID methods have obtained great progress. However, paired data are usually synthesized by hand, and the realistic blurs are more complex than synthetic ones, which makes the supervised methods inept at modeling realistic blurs and hinders real-world applications. As such, unsupervised deep BID methods without paired data offer certain advantages, but current methods still suffer from some drawbacks, e.g., bulky model size, long inference time, and strict image resolution and domain requirements. In this paper, we propose a lightweight and real-time unsupervised BID baseline, termed Frequency-domain Contrastive Loss Constrained Lightweight CycleGAN (shortly, FCL-GAN), with attractive properties, i.e., no image domain limitation, no image resolution limitation, 25x lighter than SOTA, and 5x faster than SOTA. To guarantee the lightweight property and performance superiority, two new collaboration units called lightweight domain conversion unit (LDCU) and parameter-free frequency-domain contrastive unit (PFCU) are designed. LDCU mainly implements inter-domain conversion in lightweight manner. PFCU further explores the similarity measure, external difference and internal connection between the blurred domain and sharp domain images in frequency domain, without involving extra parameters. Extensive experiments on several image datasets demonstrate the effectiveness of our FCL-GAN in terms of performance, model size and inference time.
dc.language	en
dc.publisher	Association for Computing Machinery (ACM)
dc.relation.ispartof	MM 2022 - Proceedings of the 30th ACM International Conference on Multimedia
dc.relation.ispartof	Proceedings of the 30th ACM International Conference on Multimedia
dc.relation.isbasedon	10.1145/3503161.3548113
dc.rights	info:eu-repo/semantics/closedAccess
dc.title	FCL-GAN: A Lightweight and Real-Time Baseline for Unsupervised Blind Image Deblurring
dc.type	Conference Proceeding
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
utslib.copyright.status	closed_access	*
dc.date.updated	2023-04-17T05:57:56Z
pubs.publication-status	Published

Abstract:

Blind image deblurring (BID) remains a challenging and significant task. Benefiting from the strong fitting ability of deep learning, paired data-driven supervised BID methods have obtained great progress. However, paired data are usually synthesized by hand, and the realistic blurs are more complex than synthetic ones, which makes the supervised methods inept at modeling realistic blurs and hinders real-world applications. As such, unsupervised deep BID methods without paired data offer certain advantages, but current methods still suffer from some drawbacks, e.g., bulky model size, long inference time, and strict image resolution and domain requirements. In this paper, we propose a lightweight and real-time unsupervised BID baseline, termed Frequency-domain Contrastive Loss Constrained Lightweight CycleGAN (shortly, FCL-GAN), with attractive properties, i.e., no image domain limitation, no image resolution limitation, 25x lighter than SOTA, and 5x faster than SOTA. To guarantee the lightweight property and performance superiority, two new collaboration units called lightweight domain conversion unit (LDCU) and parameter-free frequency-domain contrastive unit (PFCU) are designed. LDCU mainly implements inter-domain conversion in lightweight manner. PFCU further explores the similarity measure, external difference and internal connection between the blurred domain and sharp domain images in frequency domain, without involving extra parameters. Extensive experiments on several image datasets demonstrate the effectiveness of our FCL-GAN in terms of performance, model size and inference time.

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

http://hdl.handle.net/10453/169860