Skip to Main content Skip to Navigation
Journal articles

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

Iasonas Kokkinos 1 Rachid Deriche 2 Olivier Faugeras 2 Petros Maragos 1 
2 ODYSSEE - Computer and biological vision
DI-ENS - Département d'informatique - ENS Paris, CRISAM - Inria Sophia Antipolis - Méditerranée , ENS-PSL - École normale supérieure - Paris, Inria Paris-Rocquencourt, ENPC - École des Ponts ParisTech
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
Complete list of metadata
Contributor : Alain Monteil Connect in order to contact the contributor
Submitted on : Wednesday, November 4, 2009 - 5:33:41 PM
Last modification on : Thursday, March 17, 2022 - 10:08:29 AM


  • HAL Id : inria-00429871, version 1



Iasonas Kokkinos, Rachid Deriche, Olivier Faugeras, Petros Maragos. Computational analysis and learning for a biologically motivated model of boundary detection. Neurocomputing, Elsevier, 2008, 71 (10-12), pp.1798--1812. ⟨inria-00429871⟩



Record views