Neural Network Components for Evolved Vision Systems

Created by W.Langdon from gp-bibliography.bib Revision:1.8010

@InProceedings{Kyle-Davidson:2018:CEEC,

• author = "Cameron P. Kyle-Davidson and Adrian F. Clark",
• booktitle = "2018 10th Computer Science and Electronic Engineering (CEEC)",
• title = "Neural Network Components for Evolved Vision Systems",
• year = "2018",
• pages = "202--207",
• abstract = "This paper investigates the problem of integrating convolutional neural networks with a computer vision system that uses genetic programming to evolve image classifiers. In previous cases, genetic algorithms have been used to create and tune hyper-parameters of neural networks. This paper presents a different method of combining these two machine learning techniques, by providing trained neural networks as components that the genetic system may use. This approach makes such networks available to the genetic algorithm during training time, allowing the final evolved program to potentially leverage the pattern recognition and image classification abilities of various trained convolutional neural networks to improve final classification accuracy.",
• keywords = "genetic algorithms, genetic programming",
• DOI = "doi:10.1109/CEEC.2018.8674219",
• month = sep,
• notes = "Also known as \cite{8674219}",

}

Genetic Programming entries for Cameron P Kyle-Davidson Adrian F Clark

Citations