Please use this identifier to cite or link to this item:
https://hdl.handle.net/2440/64728
Citations | ||
Scopus | Web of Science® | Altmetric |
---|---|---|
?
|
?
|
Type: | Conference paper |
Title: | Multi-structure model selection via kernel optimisation |
Author: | Chin, T. Suter, D. Wang, H. |
Citation: | Proceedings of 2010 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2010; pp.3586-3593 |
Publisher: | IEEE COMPUTER SOC |
Publisher Place: | 10662 LOS VAQUEROS CIRCLE, PO BOX 3014, LOS ALAMITOS, CA 90720-1264 USA |
Issue Date: | 2010 |
Series/Report no.: | IEEE Conference on Computer Vision and Pattern Recognition |
ISBN: | 9781424469840 |
ISSN: | 1063-6919 |
Conference Name: | IEEE Conference on Computer Vision and Pattern Recognition (23rd : 2010 : San Francisco, CA) |
Statement of Responsibility: | Tat-Jun Chin, David Suter and Hanzi Wang |
Abstract: | Our goal is to fit the multiple instances (or structures) of a generic model existing in data. Here we propose a novel model selection scheme to estimate the number of genuine structures present. In contrast to conventional model selection approaches, our method is driven by kernel-based learning. The input data is first clustered based on their potential to have emerged from the same structure. However the number of clusters is deliberately overestimated to obtain a set of initial model fits onto the data. We then resolve the oversegmentation via a series of kernel optimisation conducted through multiple kernel learning, and the concept of kernel-target alignment is used as a model selection criterion. Experiments on synthetic and real data show that our method outperforms previous model selection schemes. We also focus on the application of multi-body motion segmentation. In particular we demonstrate success on estimating the number of motions on sequences with more than 3 unique motions. |
Rights: | ©2010 IEEE |
DOI: | 10.1109/CVPR.2010.5539931 |
Published version: | http://dx.doi.org/10.1109/cvpr.2010.5539931 |
Appears in Collections: | Aurora harvest 5 Computer Science publications |
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.