Please use this identifier to cite or link to this item: https://doi.org/10.1109/ICCV.2013.173
Title: Perspective motion segmentation via collaborative clustering
Authors: Li, Z.
Guo, J.
Cheong, L.-F. 
Zhou, S.Z. 
Issue Date: 2013
Citation: Li, Z., Guo, J., Cheong, L.-F., Zhou, S.Z. (2013). Perspective motion segmentation via collaborative clustering. Proceedings of the IEEE International Conference on Computer Vision : 1369-1376. ScholarBank@NUS Repository. https://doi.org/10.1109/ICCV.2013.173
Abstract: This paper addresses real-world challenges in the motion segmentation problem, including perspective effects, missing data, and unknown number of motions. It first formulates the 3-D motion segmentation from two perspective views as a subspace clustering problem, utilizing the epipolar constraint of an image pair. It then combines the point correspondence information across multiple image frames via a collaborative clustering step, in which tight integration is achieved via a mixed norm optimization scheme. For model selection, we propose an over-segment and merge approach, where the merging step is based on the property of the ell-1-norm of the mutual sparse representation of two over-segmented groups. The resulting algorithm can deal with incomplete trajectories and perspective effects substantially better than state-of-the-art two-frame and multi-frame methods. Experiments on a 62-clip dataset show the significant superiority of the proposed idea in both segmentation accuracy and model selection. © 2013 IEEE.
Source Title: Proceedings of the IEEE International Conference on Computer Vision
URI: http://scholarbank.nus.edu.sg/handle/10635/84080
ISBN: 9781479928392
DOI: 10.1109/ICCV.2013.173
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

SCOPUSTM   
Citations

22
checked on Aug 15, 2018

WEB OF SCIENCETM
Citations

10
checked on Jul 23, 2018

Page view(s)

26
checked on Aug 3, 2018

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.