Publications

Robustness to Adversarial Examples through an Ensemble of Specialists

Mahdieh Abbasi and Christian Gagné
En savoir plus...
Abstract - We are proposing to use an ensemble of diverse specialists, where speciality is defined according to the confusion matrix. Indeed, we observed that for adversarial instances originating from a given class, labeling tend to be done into a small subset of (incorrect) classes. Therefore, we argue that an ensemble of specialists should be better able to identify and reject fooling instances, with a high entropy (i.e., disagreement) over the decisions in the presence of adversaries. Experimental results obtained confirm that interpretation, opening a way to make the system more robust to adversarial examples through a rejection mechanism, rather than trying to classify them properly at any cost.

Bibtex:

@inproceedings{Abbasi1150,
author    = { Mahdieh Abbasi and Christian Gagné },
title     = { Robustness to Adversarial Examples through an Ensemble of Specialists },
booktitle = { International Conference on Learning Representations (ICLR), Workshop Track },
year      = { 2017 },
month     = { 4 },
journal   = { ArXiv e-prints },
web       = { https://arxiv.org/abs/1702.06856 }
}

Dernière modification: 2017/02/22 par cgagne

		Accueil \| A propos \| Personnes \| Recherche \| Publications \| Événements \| Profil
		©2002-. Laboratoire de Vision et Systèmes Numériques. Tous droits réservés