Enregistré dans:
Détails bibliographiques
Auteurs principaux: Jain, Anubhooti, Roy, Susim, Gupta, Kwanit, Vatsa, Mayank, Singh, Richa
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:https://arxiv.org/abs/2409.19619
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866929520505782272
author Jain, Anubhooti
Roy, Susim
Gupta, Kwanit
Vatsa, Mayank
Singh, Richa
author_facet Jain, Anubhooti
Roy, Susim
Gupta, Kwanit
Vatsa, Mayank
Singh, Richa
contents Deep learning models, such as those used for face recognition and attribute prediction, are susceptible to manipulations like adversarial noise and unintentional noise, including Gaussian and impulse noise. This paper introduces CIAI, a Class-Independent Adversarial Intent detection network built on a modified vision transformer with detection layers. CIAI employs a novel loss function that combines Maximum Mean Discrepancy and Center Loss to detect both intentional (adversarial attacks) and unintentional noise, regardless of the image class. It is trained in a multi-step fashion. We also introduce the aspect of intent during detection that can act as an added layer of security. We further showcase the performance of our proposed detector on CelebA, CelebA-HQ, LFW, AgeDB, and CIFAR-10 datasets. Our detector is able to detect both intentional (like FGSM, PGD, and DeepFool) and unintentional (like Gaussian and Salt & Pepper noises) perturbations.
format Preprint
id arxiv_https___arxiv_org_abs_2409_19619
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Discerning the Chaos: Detecting Adversarial Perturbations while Disentangling Intentional from Unintentional Noises
Jain, Anubhooti
Roy, Susim
Gupta, Kwanit
Vatsa, Mayank
Singh, Richa
Computer Vision and Pattern Recognition
Artificial Intelligence
Deep learning models, such as those used for face recognition and attribute prediction, are susceptible to manipulations like adversarial noise and unintentional noise, including Gaussian and impulse noise. This paper introduces CIAI, a Class-Independent Adversarial Intent detection network built on a modified vision transformer with detection layers. CIAI employs a novel loss function that combines Maximum Mean Discrepancy and Center Loss to detect both intentional (adversarial attacks) and unintentional noise, regardless of the image class. It is trained in a multi-step fashion. We also introduce the aspect of intent during detection that can act as an added layer of security. We further showcase the performance of our proposed detector on CelebA, CelebA-HQ, LFW, AgeDB, and CIFAR-10 datasets. Our detector is able to detect both intentional (like FGSM, PGD, and DeepFool) and unintentional (like Gaussian and Salt & Pepper noises) perturbations.
title Discerning the Chaos: Detecting Adversarial Perturbations while Disentangling Intentional from Unintentional Noises
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2409.19619