Computational analysis and learning for a biologically motivated model of boundary detection

Kokkinos, I; Deriche, R; Faugeras, O; Maragos, P

dc.contributor.author	Kokkinos, I	en
dc.contributor.author	Deriche, R	en
dc.contributor.author	Faugeras, O	en
dc.contributor.author	Maragos, P	en
dc.date.accessioned	2014-03-01T01:28:04Z
dc.date.available	2014-03-01T01:28:04Z
dc.date.issued	2008	en
dc.identifier.issn	0925-2312	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/18692
dc.subject	Biological vision	en
dc.subject	Boltzmann machines	en
dc.subject	Boundary detection	en
dc.subject	Computer vision	en
dc.subject	Learning	en
dc.subject	Mean field theory	en
dc.subject	Neural networks	en
dc.subject	Perceptual grouping	en
dc.subject	Variational techniques	en
dc.subject.classification	Computer Science, Artificial Intelligence	en
dc.subject.other	Algorithms	en
dc.subject.other	Cost functions	en
dc.subject.other	Image segmentation	en
dc.subject.other	Mean field theory	en
dc.subject.other	Neural networks	en
dc.subject.other	Variational techniques	en
dc.subject.other	Biological vision	en
dc.subject.other	Boltzmann machines	en
dc.subject.other	Boundary detection	en
dc.subject.other	Perceptual grouping	en
dc.subject.other	Computer vision	en
dc.subject.other	algorithm	en
dc.subject.other	article	en
dc.subject.other	computer network	en
dc.subject.other	computer system	en
dc.subject.other	image display	en
dc.subject.other	image processing	en
dc.subject.other	intermethod comparison	en
dc.subject.other	machine learning	en
dc.subject.other	mathematical computing	en
dc.subject.other	mathematical model	en
dc.subject.other	priority journal	en
dc.subject.other	quality control	en
dc.subject.other	visual information	en
dc.title	Computational analysis and learning for a biologically motivated model of boundary detection	en
heal.type	journalArticle	en
heal.identifier.primary	10.1016/j.neucom.2007.11.031	en
heal.identifier.secondary	http://dx.doi.org/10.1016/j.neucom.2007.11.031	en
heal.language	English	en
heal.publicationDate	2008	en
heal.abstract	In this work we address the problem of boundary detection by combining ideas and approaches from biological and computational vision. Initially, we propose a simple and efficient architecture that is inspired from models of biological vision. Subsequently, we interpret and learn the system using computer vision techniques: First, we present analogies between the system components and computer vision techniques and interpret the network as minimizing a cost functional, thereby establishing a link with variational techniques. Second, based on Mean Field Theory the equations describing the network behavior are interpreted statistically. Third, we build on this interpretation to develop an algorithm to learn the network weights from manually segmented natural images. Using a systematic evaluation on the Berkeley benchmark we show that when using the learned connection weights our network outperforms classical edge detection algorithms. (C) 2008 Elsevier B.V. All rights reserved.	en
heal.publisher	ELSEVIER SCIENCE BV	en
heal.journalName	Neurocomputing	en
dc.identifier.doi	10.1016/j.neucom.2007.11.031	en
dc.identifier.isi	ISI:000257413300004	en
dc.identifier.volume	71	en
dc.identifier.issue	10-12	en
dc.identifier.spage	1798	en
dc.identifier.epage	1812	en