Unsupervised Point Cloud Pre-Training Via Contrasting and Clustering

Mei, G; Huang, X; Liu, J; Zhang, J; Wu, Q

Unsupervised Point Cloud Pre-Training Via Contrasting and Clustering

Mei, G

Huang, X Liu, J Zhang, J Wu, Q

Permalink

Publisher:: Institute of Electrical and Electronics Engineers (IEEE)
Publication Type:: Conference Proceeding
Citation:: Proceedings - International Conference on Image Processing, ICIP, 2022, 00, pp. 66-70
Issue Date:: 2022-10-16

Embargoed

	Filename	Description	Size
	Binder2.pdf	Accepted version	429.49 kB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Embargoed
Open Access

This item is currently unavailable due to the publisher's embargo.

The embargo period expires on 16 Oct 2024

Full metadata record

Field	Value	Language
dc.contributor.author	Mei, G https://orcid.org/0000-0002-0494-5031
dc.contributor.author	Huang, X
dc.contributor.author	Liu, J
dc.contributor.author	Zhang, J
dc.contributor.author	Wu, Q https://orcid.org/0000-0001-5641-2483
dc.date	2022-10-16
dc.date.accessioned	2023-04-12T23:39:31Z
dc.date.available	2023-04-12T23:39:31Z
dc.date.issued	2022-10-16
dc.identifier.citation	Proceedings - International Conference on Image Processing, ICIP, 2022, 00, pp. 66-70
dc.identifier.isbn	9781665496209
dc.identifier.issn	1522-4880
dc.identifier.uri	http://hdl.handle.net/10453/169685
dc.description.abstract	The annotation for large-scale point clouds is still time-consuming and unavailable for many complex real-world tasks. Point cloud pre-training is a promising direction to auto-extract features without labeled data. Therefore, this paper proposes a general unsupervised approach, named ConClu for point cloud pre-training by jointly performing contrasting and clustering. Specifically, the contrasting is formulated by maximizing the similarity feature vectors produced by encoders fed with two augmentations of the same point cloud. The clustering simultaneously clusters the data while enforcing consistency between cluster assignments produced different augmentations. Experimental evaluations on downstream applications outperform state-of-the-art techniques, which demonstrates the effectiveness of our framework.
dc.language	en
dc.publisher	Institute of Electrical and Electronics Engineers (IEEE)
dc.relation.ispartof	Proceedings - International Conference on Image Processing, ICIP
dc.relation.ispartof	2022 IEEE International Conference on Image Processing (ICIP)
dc.relation.isbasedon	10.1109/icip46576.2022.9897388
dc.rights	info:eu-repo/semantics/embargoedAccess
dc.title	Unsupervised Point Cloud Pre-Training Via Contrasting and Clustering
dc.type	Conference Proceeding
utslib.citation.volume	00
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 - INEXT - Innovation in IT Services and Applications
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	embargoed	*
utslib.copyright.embargo	2024-10-16T00:00:00+1000Z
dc.date.updated	2023-04-12T23:39:30Z
pubs.finish-date	2022-10-19
pubs.publication-status	Published
pubs.start-date	2022-10-16
pubs.volume	00

Abstract:

The annotation for large-scale point clouds is still time-consuming and unavailable for many complex real-world tasks. Point cloud pre-training is a promising direction to auto-extract features without labeled data. Therefore, this paper proposes a general unsupervised approach, named ConClu for point cloud pre-training by jointly performing contrasting and clustering. Specifically, the contrasting is formulated by maximizing the similarity feature vectors produced by encoders fed with two augmentations of the same point cloud. The clustering simultaneously clusters the data while enforcing consistency between cluster assignments produced different augmentations. Experimental evaluations on downstream applications outperform state-of-the-art techniques, which demonstrates the effectiveness of our framework.

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

http://hdl.handle.net/10453/169685