Accession Number : ADA264800


Title :   Principal Components of Natural Images: An Analytical Solution


Descriptive Note : Technical rept.,


Corporate Author : BROWN UNIV PROVIDENCE RI INST FOR BRAIN AND NEURAL SYSTEMS


Personal Author(s) : Liu, Yong ; Shouval, Harel


Full Text : https://apps.dtic.mil/dtic/tr/fulltext/u2/a264800.pdf


Report Date : 17 May 1993


Pagination or Media Count : 9


Abstract : The structure of receptive fields in the visual cortex is believed to be shaped by unsupervised learning. A simple variant of unsupervised learning is the extraction of principal components. In this paper, we derived analytically the form of the principal components of natural images. This derivation relies on results about the covariance matrix of natural images. Our results predict both the shapes and the phases of the receptive fields. We also compared our results to numerical simulation results. Finally the biological relevance of our results is discussed.


Descriptors :   *VISION , *LEARNING , *RECEPTOR SITES(PHYSIOLOGY) , *VISUAL CORTEX , COMPUTERIZED SIMULATION , IMAGE PROCESSING , NUMERICAL METHODS AND PROCEDURES , COVARIANCE , MATRICES(MATHEMATICS)


Subject Categories : Anatomy and Physiology


Distribution Statement : APPROVED FOR PUBLIC RELEASE