Home > Research > Publications & Outputs > A cascade of deep learning fuzzy rule-based ima...

Electronic data

  • SMC_MICEv2.0

    Rights statement: ©2017 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

    Accepted author manuscript, 648 KB, PDF document

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License

Links

Text available via DOI:

View graph of relations

A cascade of deep learning fuzzy rule-based image classifier and SVM

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paper

Published
Publication date5/10/2017
Host publicationSystems, Man, and Cybernetics (SMC), 2017 IEEE International Conference on: Human Intelligence for Systems and Cybernetics
PublisherIEEE
Pages746-751
Number of pages6
ISBN (Electronic)9781538616451
ISBN (Print)9781538616468
Original languageEnglish
EventIEEE SMC conference -
Duration: 5/10/2017 → …

Conference

ConferenceIEEE SMC conference
Period5/10/17 → …

Conference

ConferenceIEEE SMC conference
Period5/10/17 → …

Abstract

In this paper, a fast, transparent, self-evolving, deep learning fuzzy rule-based (DLFRB) image classifier is proposed. This new classifier is a cascade of the recently introduced DLFRB classifier and a SVM based auxiliary. The DLFRB classifier serves as the main engine and can identify a number of human interpretable fuzzy rules through a very short, transparent, highly parallelizable training process. The SVM based auxiliary plays the role as a conflict resolver when the DLFRB classifier produces two highly confident labels for a single image. Only the fundamental image transformation techniques (rotation, scaling and segmentation) and feature descriptors (GIST and HOG) are used for pre-processing and feature extraction, but the proposed approach significantly outperforms the state-of-art methods in terms of both time and precision. Numerical experiments based on a handwriting digits recognition problem are used to demonstrate the highly accurate and repeatable performance of the proposed approach after a very shorting training process.

Bibliographic note

©2017 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.