Eliminating Useless Object Detectors Evolved in Multiple-Objective Genetic Programming

Created by W.Langdon from gp-bibliography.bib Revision:1.3973

@InProceedings{conf/ausai/ScobleJZ11,
  author =       "Aaron Scoble and Mark Johnston and Mengjie Zhang",
  title =        "Eliminating Useless Object Detectors Evolved in
                 Multiple-Objective Genetic Programming",
  booktitle =    "Proceedings of the 24th Australasian Joint Conference
                 Advances in Artificial Intelligence (AI 2011)",
  year =         "2011",
  editor =       "Dianhui Wang and Mark Reynolds",
  volume =       "7106",
  series =       "Lecture Notes in Computer Science",
  pages =        "341--350",
  address =      "Perth, Australia",
  month =        dec # " 5-8",
  publisher =    "Springer",
  keywords =     "genetic algorithms, genetic programming",
  isbn13 =       "978-3-642-25831-2",
  DOI =          "doi:10.1007/978-3-642-25832-9_35",
  size =         "10 pages",
  abstract =     "Object detection is the task of correctly identifying
                 and locating objects of interest within a larger image.
                 An ideal object detector would maximise the number of
                 correctly located objects and minimise the number of
                 false-alarms. Previous work, following the traditional
                 multiple-objective paradigm of finding Pareto-optimal
                 tradeoffs between these objectives, suffers from an
                 abundance of useless detectors that either detect
                 nothing (but with no false-alarms) or mark every pixel
                 as an object (perfect detection performance with but a
                 very large number of false-alarms); these are very
                 often Pareto-optimal and hence inadvertently rewarded.
                 We propose and compare a number of improvements to
                 eliminate useless detectors during evolution. The most
                 successful improvements are generally more inefficient
                 than the benchmark MOGP approach due to the often vast
                 numbers of additional crossover and mutation operations
                 required, but as a result the archive populations
                 generally include a much higher number of
                 Pareto-fronts.",
  affiliation =  "School of Mathematics, Statistics and Operations
                 Research, Victoria University of Wellington, PO Box
                 600, Wellington, New Zealand",
  bibdate =      "2011-12-02",
  bibsource =    "DBLP,
                 http://dblp.uni-trier.de/db/conf/ausai/ausai2011.html#ScobleJZ11",
}

Genetic Programming entries for Aaron Scoble Mark Johnston Mengjie Zhang

Citations