Efficient dense subspace clustering

Files

RA_hdl_108963.pdf (859.13 KB)
  (Restricted Access)

Date

2014

Authors

Ji, P.
Salzamnn, M.
Li, H.

Editors

Advisors

Journal Title

Journal ISSN

Volume Title

Type:

Conference paper

Citation

Proceedings / IEEE Workshop on Applications of Computer Vision. IEEE Workshop on Applications of Computer Vision, 2014, pp.461-468

Statement of Responsibility

Pan Ji, Mathieu Salzmann, Hongdong Li

Conference Name

2014 IEEE Winter Conference on Applications of Computer Vision (WACV 2014) (24 Mar 2014 - 26 Mar 2014 : Steamboat Springs, CO, USA)

Abstract

In this paper, we tackle the problem of clustering data points drawn from a union of linear (or affine) subspaces. To this end, we introduce an efficient subspace clustering algorithm that estimates dense connections between the points lying in the same subspace. In particular, instead of following the standard compressive sensing approach, we formulate subspace clustering as a Frobenius norm minimization problem, which inherently yields denser con- nections between the data points. While in the noise-free case we rely on the self-expressiveness of the observations, in the presence of noise we simultaneously learn a clean dictionary to represent the data. Our formulation lets us address the subspace clustering problem efficiently. More specifically, the solution can be obtained in closed-form for outlier-free observations, and by performing a series of linear operations in the presence of outliers. Interestingly, we show that our Frobenius norm formulation shares the same solution as the popular nuclear norm minimization approach when the data is free of any noise, or, in the case of corrupted data, when a clean dictionary is learned. Our experimental evaluation on motion segmentation and face clustering demonstrates the benefits of our algorithm in terms of clustering accuracy and efficiency.

School/Discipline

Dissertation Note

Provenance

Description

Access Status

Rights

Copyright © 2014, IEEE

License

Grant ID

Call number

Persistent link to this record