Home // PATTERNS 2018, The Tenth International Conference on Pervasive Patterns and Applications // View article

Automatic Construction of Large Scale Image Data Set from Web Using Ontology and Deep Learning Model

Authors:
Takahiro Yoshimura
Yuki Mizoguchi
Yuji Iwahori
Ryosuke Yamanishi

Keywords: Image data set; Web image mining; Image processing; Web intelligence.

Abstract:
General object recognition requires a large scale image data set and recognition accuracy depends on the image data set used for the learning. This paper proposes a method to collect only the target images using Ontology and features obtained from obtained from Convolutional Neural Network (CNN). The proposed approach constructs large scale data set automatically by expanding the range of collected images. It is shown that the proposed approach is effective by collecting the image data set and a mean accuracy 88.6% was obtained from the subjective evaluation in experiments.

Pages: 21 to 22

Copyright: Copyright (c) IARIA, 2018

Publication date: February 18, 2018

Published in: conference

ISSN: 2308-3557

ISBN: 978-1-61208-612-5

Location: Barcelona, Spain

Dates: from February 18, 2018 to February 22, 2018