Share Email Print

Proceedings Paper

Learning efficient linear codes for natural images: the roles of sparseness, overcompleteness, and statistical independence
Author(s): Bruno A. Olshausen; David J. Field
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

An algorithm is described which allows for the learning of sparse, overcomplete image representations. Images are modeled as a linear superposition of basis functions, and a set of basis functions is sought which maximizes the sparseness of the representation (fewest number of active units per image). When applied to natural scenes, the basis functions converge to localized, oriented, bandpass functions that bear a strong resemblance to the receptive fields of neurons in the primate striate cortex. Importantly, the code can be made overcomplete, which allows for an increased degree of sparseness in which the basis functions can become more specialized. The learned basis functions constitute an efficient representation of natural images because sparseness forces a form of reduced entropy representation that minimizes statistical dependencies among outputs.

Paper Details

Date Published: 22 April 1996
PDF: 7 pages
Proc. SPIE 2657, Human Vision and Electronic Imaging, (22 April 1996); doi: 10.1117/12.238708
Show Author Affiliations
Bruno A. Olshausen, Cornell Univ. (United States)
David J. Field, Cornell Univ. (United States)

Published in SPIE Proceedings Vol. 2657:
Human Vision and Electronic Imaging
Bernice E. Rogowitz; Jan P. Allebach, Editor(s)

© SPIE. Terms of Use
Back to Top