An Overcomplete ICA Algorithm by InfoMax and InfoMin

Authors:
Yoshitatsu Matsuda;Kazunori Yamaguchi
Affiliations:
Department of Integrated Information Technology, Aoyama Gakuin University, Kanagawa, Japan 229-8558;Department of General Systems Studies, Graduate School of Arts and Sciences, The University of Tokyo, Tokyo, Japan 153-8902
Venue:
ICANN '08 Proceedings of the 18th international conference on Artificial Neural Networks, Part I
Year:
2008

Citing 7
Cited 0

Self-Organization in a Perceptual Network

Computer
An information-maximization approach to blind separation and blind deconvolution

Neural Computation
Estimating Overcomplete Independent Component Bases for Image Windows

Journal of Mathematical Imaging and Vision
Topographic Independent Component Analysis

Neural Computation
Learning Overcomplete Representations

Neural Computation
A fixed-point algorithm of topographic ICA

ICANN'06 Proceedings of the 16th international conference on Artificial Neural Networks - Volume Part II
The infomin principle for ICA and topographic mappings

ICA'06 Proceedings of the 6th international conference on Independent Component Analysis and Blind Signal Separation

Quantified Score

Hi-index	0.00

Visualization

Abstract

It is known that the number of the edge detectors significantly exceeds that of input signals in the visual system of the brains. This phenomenon has been often regarded as overcomplete independent component analysis (ICA) and some generative models have been proposed. Though the models are effective, they need to assume some ad-hoc prior probabilistic models. Recently, the InfoMin principle was proposed as a comprehensive framework with minimal prior assumptions for explaining the information processing in the brains and its usefulness has been verified in the classic non-overcomplete cases. In this paper, we propose a new ICA contrast function for overcomplete cases, which is deductively derived from the the InfoMin and InfoMax principles without any prior models. Besides, we construct an efficient fixed-point algorithm for optimizing it by an approximate Newton's method. Numerical experiments verify the effectiveness of the proposed method.