Empirical Inference

Unsupervised learning of a steerable basis for invariant image representations

2007

Conference Paper

ei


There are two aspects to unsupervised learning of invariant representations of images: First, we can reduce the dimensionality of the representation by finding an optimal trade-off between temporal stability and informativeness. We show that the answer to this optimization problem is generally not unique so that there is still considerable freedom in choosing a suitable basis. Which of the many optimal representations should be selected? Here, we focus on this second aspect, and seek to find representations that are invariant under geometrical transformations occuring in sequences of natural images. We utilize ideas of steerability and Lie groups, which have been developed in the context of filter design. In particular, we show how an anti-symmetric version of canonical correlation analysis can be used to learn a full-rank image basis which is steerable with respect to rotations. We provide a geometric interpretation of this algorithm by showing that it finds the two-dimensional eigensubspaces of the avera ge bivector. For data which exhibits a variety of transformations, we develop a bivector clustering algorithm, which we use to learn a basis of generalized quadrature pairs (i.e. complex cells) from sequences of natural images.

Author(s): Bethge, M. and Gerwinn, S. and Macke, JH.
Book Title: Human Vision and Electronic Imaging XII
Journal: Human Vision and Electronic Imaging XII: Proceedings of the SPIE Human Vision and Electronic Imaging Conference 2007
Pages: 1-12
Year: 2007
Month: February
Day: 0
Editors: Rogowitz, B. E.
Publisher: SPIE

Department(s): Empirical Inference
Bibtex Type: Conference Paper (inproceedings)

DOI: 10.1117/12.711119
Event Name: SPIE Human Vision and Electronic Imaging Conference 2007
Event Place: San Jose, CA, USA

Address: Bellingham, WA, USA
Digital: 0
Language: en
Organization: Max-Planck-Gesellschaft
School: Biologische Kybernetik

Links: PDF
Web

BibTex

@inproceedings{4305,
  title = {Unsupervised learning of a steerable basis for invariant image representations},
  author = {Bethge, M. and Gerwinn, S. and Macke, JH.},
  journal = {Human Vision and Electronic Imaging XII: Proceedings of the SPIE Human Vision and Electronic Imaging Conference 2007},
  booktitle = {Human Vision and Electronic Imaging XII},
  pages = {1-12},
  editors = {Rogowitz, B. E.},
  publisher = {SPIE},
  organization = {Max-Planck-Gesellschaft},
  school = {Biologische Kybernetik},
  address = {Bellingham, WA, USA},
  month = feb,
  year = {2007},
  doi = {10.1117/12.711119},
  month_numeric = {2}
}