Adversarial purification with score-based generative models

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 60
  • Download : 0
DC FieldValueLanguage
dc.contributor.authorYoon, Jongminko
dc.contributor.authorHwang, Sung Juko
dc.contributor.authorLee, Juhoko
dc.identifier.citationThirty-eighth International Conference on Machine Learning 2021-
dc.description.abstractWhile adversarial training is considered as a standard defense method against adversarial attacks for image classifiers, adversarial purification, which purifies attacked images into clean images with a standalone purification model, has shown promises as an alternative defense method. Recently, an Energy-Based Model (EBM) trained with Markov-Chain Monte-Carlo (MCMC) has been highlighted as a purification model, where an attacked image is purified by running a long Markov-chain using the gradients of the EBM. Yet, the practicality of the adversarial purification using an EBM remains questionable because the number of MCMC steps required for such purification is too large. In this paper, we propose a novel adversarial purification method based on an EBM trained with Denoising Score-Matching (DSM). We show that an EBM trained with DSM can quickly purify attacked images within a few steps. We further introduce a simple yet effective randomized purification scheme that injects random noises into images before purification. This process screens the adversarial perturbations imposed on images by the random noises and brings the images to the regime where the EBM can denoise well. We show that our purification method is robust against various attacks and demonstrate its state-of-the-art performances.-
dc.publisherInternational Conference on Machine Learning Organizing Committee-
dc.titleAdversarial purification with score-based generative models-
dc.citation.publicationnameThirty-eighth International Conference on Machine Learning 2021-
dc.contributor.localauthorHwang, Sung Ju-
dc.contributor.localauthorLee, Juho-
dc.contributor.nonIdAuthorYoon, Jongmin-
Appears in Collection
RIMS Conference Papers
Files in This Item
There are no files associated with this item.


  • mendeley


rss_1.0 rss_2.0 atom_1.0