Evolution of Visual Feature Detectors

Created by W.Langdon from gp-bibliography.bib Revision:1.4221

  author =       "Tony Belpaeme",
  title =        "Evolution of Visual Feature Detectors",
  booktitle =    "Late Breaking Papers at EvoISAP'99: the First European
                 Workshop on Evolutionary Computation in Image Analysis
                 and Signal Processing",
  year =         "1999",
  editor =       "Riccardo Poli and Stefano Cagnoni and 
                 Hans-Michael Voigt and Terry Fogarty and Peter Nordin",
  pages =        "1--10",
  address =      "Goteborg, Sweden",
  month =        "28 " # may,
  organisation = "EvoNet",
  keywords =     "genetic algorithms, genetic programming",
  URL =          "http://arti.vub.ac.be/~tony/papers/EvoIASP99.ps.gz",
  URL =          "http://citeseer.ist.psu.edu/362631.html",
  abstract =     "This paper describes how sets of visual feature
                 detectors are evolved starting from simple primitives.
                 The primitives, of which some are inspired on visual
                 processing observed in mammalian visual pathways, are
                 combined using genetic programming to form a
                 feed-forward feature-extraction hierarchy. Input to the
                 feature detectors consists of a series of real-world
                 images, containing objects or faces. The results show
                 how each set of feature detectors self-organizes into a
                 set which is capable of returning feature vectors for
                 discriminating the input images. We discuss the
                 influence of different settings on the evolution of the
                 feature detectors and explain some phenomena.",
  notes =        "EvoIASP'99 Available as CSRP-99-10 from the School of
                 Computer Science, University of Birmingham, Edgbaston,
                 Birmingham B15 2TT, UK.

                 STGP. Information returned by each (of 5) feature
                 detector, entropy of the output vector p4 {"}if the
                 outputs are weel spread, meaning the feature detectors
                 return useful information, the fitness will be high.
                 Explicit parsimony preasure, but not needed p8?

Genetic Programming entries for Tony Belpaeme