This paper introduces a novel method of visual
learning based on Genetic Programming, which evolves a
population of individuals (image analysis programs)
that process attributed visual primitives derived from
raw raster images. The goal is to evolve an image
analysis program that correctly recognises the training
concept (shape). The approach uses generative
evaluation scheme: individuals are rewarded for
re-producing the shape of the object being recognised
using graphical primitives and elementary background
knowledge encoded in predefined operators. Evolutionary
run is driven by a multiobjective fitness function to
prevent premature convergence and enable effective
exploration of the space of solutions. We present the
method in detail and verify it experimentally on the
task of learning two visual concepts from examples.
%0 Journal Article
%1 Krawiec07PRL
%A Krawiec, Krzysztof
%D 2007
%J Pattern Recognition Letters
%K Evolutionary Generative Visual algorithms, genetic learning, of pattern programming, recognition recognition, synthesis systems
%N 16
%P 2385--2400
%R doi:10.1016/j.patrec.2007.08.001
%T Generative Learning of Visual Concepts using
Multiobjective Genetic Programming
%V 28
%X This paper introduces a novel method of visual
learning based on Genetic Programming, which evolves a
population of individuals (image analysis programs)
that process attributed visual primitives derived from
raw raster images. The goal is to evolve an image
analysis program that correctly recognises the training
concept (shape). The approach uses generative
evaluation scheme: individuals are rewarded for
re-producing the shape of the object being recognised
using graphical primitives and elementary background
knowledge encoded in predefined operators. Evolutionary
run is driven by a multiobjective fitness function to
prevent premature convergence and enable effective
exploration of the space of solutions. We present the
method in detail and verify it experimentally on the
task of learning two visual concepts from examples.
@article{Krawiec07PRL,
abstract = {This paper introduces a novel method of visual
learning based on Genetic Programming, which evolves a
population of individuals (image analysis programs)
that process attributed visual primitives derived from
raw raster images. The goal is to evolve an image
analysis program that correctly recognises the training
concept (shape). The approach uses generative
evaluation scheme: individuals are rewarded for
re-producing the shape of the object being recognised
using graphical primitives and elementary background
knowledge encoded in predefined operators. Evolutionary
run is driven by a multiobjective fitness function to
prevent premature convergence and enable effective
exploration of the space of solutions. We present the
method in detail and verify it experimentally on the
task of learning two visual concepts from examples.},
added-at = {2008-06-19T17:35:00.000+0200},
author = {Krawiec, Krzysztof},
biburl = {https://www.bibsonomy.org/bibtex/2001484a0e7787faea1d3b816f89b7aa5/brazovayeye},
doi = {doi:10.1016/j.patrec.2007.08.001},
email = {krawiec@cs.put.poznan.pl},
interhash = {2fc07df1af55889c36fe98df873a2022},
intrahash = {001484a0e7787faea1d3b816f89b7aa5},
journal = {Pattern Recognition Letters},
keywords = {Evolutionary Generative Visual algorithms, genetic learning, of pattern programming, recognition recognition, synthesis systems},
month = {1 December},
number = 16,
pages = {2385--2400},
timestamp = {2008-06-19T17:44:24.000+0200},
title = {Generative Learning of Visual Concepts using
Multiobjective Genetic Programming},
volume = 28,
year = 2007
}