Multi-objective neural network optimization for visual object detection

Stefan Roth*, Alexander Gepperth, Christian Igel

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingBook chapterResearchpeer-review

15 Citations (Scopus)

Abstract

In real-time computer vision, there is a need for classifiers that detect patterns fast and reliably. We apply multi-objective optimization (MOO) to the design of feed-forward neural networks for real-world object recognition tasks, where computational complexity and accuracy define partially conflicting objectives. Evolutionary structure optimization and pruning are compared for the adaptation of the network topology. In addition, the results of MOO are contrasted to those of a single-objective evolutionary algorithm. As a part of the evolutionary algorithm, the automatic adaptation of operator probabilities in MOO is described.

Original languageEnglish
Title of host publicationMulti-objective machine learning
EditorsYaochu Jin
Number of pages27
VolumeV
Publication date2006
Pages629-655
ISBN (Print)978-3-540-30676-4
ISBN (Electronic)978-3-540-33019-6
DOIs
Publication statusPublished - 2006
Externally publishedYes
SeriesStudies in Computational Intelligence
Volume16
ISSN1860-949X

Cite this