Mason Archival Repository Service

Diverse Knowledge Distillation (DKD): A Solution for Improving The Robustness of Ensemble Models Against Adversarial Attacks

Show simple item record

dc.contributor.author Mirzaeian, Ali
dc.contributor.author Kosecka, Jana
dc.contributor.author Homayoun, Houman
dc.contributor.author Mohsenin, Tinoosh
dc.contributor.author Sasan, Avesta
dc.date.accessioned 2021-02-03T19:21:55Z
dc.date.available 2021-02-03T19:21:55Z
dc.date.issued 2021
dc.identifier.citation Ali Mirzaeian, Jana Kosecka, Houman Homayoun, Tinoosh Mohsenin, Avesta Sasan. Diverse Knowledge Distillation (DKD): A Solution for Improving The Robustness of Ensemble Models Against Adversarial Attacks. California, ISQED 2021 en_US
dc.identifier.uri http://hdl.handle.net/1920/11947
dc.description.abstract This paper proposes an ensemble learning model that is resistant to adversarial attacks. To build resilience, we introduced a training process where each member learns a radically distinct latent space. Member models are added one at a time to the ensemble. Simultaneously, the loss function is regulated by a reverse knowledge distillation, forcing the new member to learn different features and map to a latent space safely distanced from those of existing members. We assessed the security and performance of the proposed solution on image classification tasks using CIFAR10 and MNIST datasets and showed security and performance improvement compared to the state of the art defense methods. en_US
dc.language.iso en_US en_US
dc.rights Attribution-ShareAlike 3.0 United States *
dc.rights.uri http://creativecommons.org/licenses/by-sa/3.0/us/ *
dc.subject Ensemble Learning en_US
dc.subject Neural Networks en_US
dc.title Diverse Knowledge Distillation (DKD): A Solution for Improving The Robustness of Ensemble Models Against Adversarial Attacks en_US
dc.type Working Paper en_US


Files in this item

The following license files are associated with this item:

This item appears in the following Collection(s)

Show simple item record

Attribution-ShareAlike 3.0 United States Except where otherwise noted, this item's license is described as Attribution-ShareAlike 3.0 United States

Search MARS


Browse

My Account

Statistics