Fusion of Evidential CNN Classifiers for Image Classification

Zheng Tong, Philippe Xu, Thierry Denœux

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We propose an information-fusion approach based on belief functions to combine convolutional neural networks. In this approach, several pre-trained DS-based CNN architectures extract features from input images and convert them into mass functions on different frames of discernment. A fusion module then aggregates these mass functions using Dempster’s rule. An end-to-end learning procedure allows us to fine-tune the overall architecture using a learning set with soft labels, which further improves the classification performance. The effectiveness of this approach is demonstrated experimentally using three benchmark databases.

Original languageEnglish
Title of host publicationBelief Functions
Subtitle of host publicationTheory and Applications - 6th International Conference, BELIEF 2021, Proceedings
EditorsThierry Denœux, Eric Lefèvre, Zhunga Liu, Frédéric Pichon
PublisherSpringer Science and Business Media Deutschland GmbH
Pages168-176
Number of pages9
ISBN (Print)9783030886004
DOIs
Publication statusPublished - 1 Jan 2021
Externally publishedYes
Event6th International Conference on Belief Functions, BELIEF 2021 - Virtual, Online
Duration: 15 Oct 202119 Oct 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12915 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference6th International Conference on Belief Functions, BELIEF 2021
CityVirtual, Online
Period15/10/2119/10/21

Keywords

  • Convolutional neural network
  • Dempster-Shafer theory
  • Evidence theory
  • Information fusion
  • Object recognition

Fingerprint

Dive into the research topics of 'Fusion of Evidential CNN Classifiers for Image Classification'. Together they form a unique fingerprint.

Cite this