QuEst: Adversarial Attack Intensity Estimation via Query Response Analysis

Deep learning has dramatically advanced computer vision tasks, including person re-identification (re-ID), substantially improving matching individuals across diverse camera views. However, person re-ID systems remain vulnerable to adversarial attacks that introduce imperceptible perturbations, lead...

Full description

Saved in:
Bibliographic Details
Main Authors: Eun Gi Lee, Chi Hyeok Min, Seok Bong Yoo
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Mathematics
Subjects:
Online Access:https://www.mdpi.com/2227-7390/12/22/3508
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846153056760102912
author Eun Gi Lee
Chi Hyeok Min
Seok Bong Yoo
author_facet Eun Gi Lee
Chi Hyeok Min
Seok Bong Yoo
author_sort Eun Gi Lee
collection DOAJ
description Deep learning has dramatically advanced computer vision tasks, including person re-identification (re-ID), substantially improving matching individuals across diverse camera views. However, person re-ID systems remain vulnerable to adversarial attacks that introduce imperceptible perturbations, leading to misidentification and undermining system reliability. This paper addresses the challenge of robust person re-ID in the presence of adversarial examples by estimating attack intensity to enable effective detection and adaptive purification. The proposed approach leverages the observation that adversarial examples in retrieval tasks disrupt the relevance and internal consistency of retrieval results, degrading re-ID accuracy. This approach estimates the attack intensity and dynamically adjusts the purification strength by analyzing the query response data, addressing the limitations of fixed purification methods. This approach also preserves the performance of the model on clean data by avoiding unnecessary manipulation while improving the robustness of the system and its reliability in the presence of adversarial examples. The experimental results demonstrate that the proposed method effectively detects adversarial examples and estimates the attack intensity through query response analysis. This approach enhances purification performance when integrated with adversarial purification techniques in person re-ID systems.
format Article
id doaj-art-32c44cd075a846c2bd6e346df8be260f
institution Kabale University
issn 2227-7390
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Mathematics
spelling doaj-art-32c44cd075a846c2bd6e346df8be260f2024-11-26T18:11:39ZengMDPI AGMathematics2227-73902024-11-011222350810.3390/math12223508QuEst: Adversarial Attack Intensity Estimation via Query Response AnalysisEun Gi Lee0Chi Hyeok Min1Seok Bong Yoo2Department of Artificial Intelligence Convergence, Chonnam National University, Gwangju 61186, Republic of KoreaDepartment of Artificial Intelligence Convergence, Chonnam National University, Gwangju 61186, Republic of KoreaDepartment of Artificial Intelligence Convergence, Chonnam National University, Gwangju 61186, Republic of KoreaDeep learning has dramatically advanced computer vision tasks, including person re-identification (re-ID), substantially improving matching individuals across diverse camera views. However, person re-ID systems remain vulnerable to adversarial attacks that introduce imperceptible perturbations, leading to misidentification and undermining system reliability. This paper addresses the challenge of robust person re-ID in the presence of adversarial examples by estimating attack intensity to enable effective detection and adaptive purification. The proposed approach leverages the observation that adversarial examples in retrieval tasks disrupt the relevance and internal consistency of retrieval results, degrading re-ID accuracy. This approach estimates the attack intensity and dynamically adjusts the purification strength by analyzing the query response data, addressing the limitations of fixed purification methods. This approach also preserves the performance of the model on clean data by avoiding unnecessary manipulation while improving the robustness of the system and its reliability in the presence of adversarial examples. The experimental results demonstrate that the proposed method effectively detects adversarial examples and estimates the attack intensity through query response analysis. This approach enhances purification performance when integrated with adversarial purification techniques in person re-ID systems.https://www.mdpi.com/2227-7390/12/22/3508computer visiondeep learningadversarial metric attackperson re-identificationattack intensity estimation
spellingShingle Eun Gi Lee
Chi Hyeok Min
Seok Bong Yoo
QuEst: Adversarial Attack Intensity Estimation via Query Response Analysis
Mathematics
computer vision
deep learning
adversarial metric attack
person re-identification
attack intensity estimation
title QuEst: Adversarial Attack Intensity Estimation via Query Response Analysis
title_full QuEst: Adversarial Attack Intensity Estimation via Query Response Analysis
title_fullStr QuEst: Adversarial Attack Intensity Estimation via Query Response Analysis
title_full_unstemmed QuEst: Adversarial Attack Intensity Estimation via Query Response Analysis
title_short QuEst: Adversarial Attack Intensity Estimation via Query Response Analysis
title_sort quest adversarial attack intensity estimation via query response analysis
topic computer vision
deep learning
adversarial metric attack
person re-identification
attack intensity estimation
url https://www.mdpi.com/2227-7390/12/22/3508
work_keys_str_mv AT eungilee questadversarialattackintensityestimationviaqueryresponseanalysis
AT chihyeokmin questadversarialattackintensityestimationviaqueryresponseanalysis
AT seokbongyoo questadversarialattackintensityestimationviaqueryresponseanalysis