@inproceedings{e5d5836936f3483383d10651fd0f03e3,
title = "Fusion of Evidential CNN Classifiers for Image Classification",
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{\textquoteright}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.",
keywords = "Convolutional neural network, Dempster-Shafer theory, Evidence theory, Information fusion, Object recognition",
author = "Zheng Tong and Philippe Xu and Thierry Den{\oe}ux",
note = "Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG.; 6th International Conference on Belief Functions, BELIEF 2021 ; Conference date: 15-10-2021 Through 19-10-2021",
year = "2021",
month = jan,
day = "1",
doi = "10.1007/978-3-030-88601-1\_17",
language = "English",
isbn = "9783030886004",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Science and Business Media Deutschland GmbH",
pages = "168--176",
editor = "Thierry Den{\oe}ux and Eric Lef{\`e}vre and Zhunga Liu and Fr{\'e}d{\'e}ric Pichon",
booktitle = "Belief Functions",
}