ADBM: Adversarial diffusion bridge model for reliable adversarial purification

Recently Diffusion-based Purification (DiffPure) has been recognized as an effective defense method against adversarial examples. However, we find DiffPure which directly employs the original pre-trained diffusion models for adversarial purification, to be suboptimal. This is due to an inherent trad...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Li, Xiao, Sun, Wenxuan, Chen, Huanran, Li, Qiongxiu, Liu, Yining, He, Yingzhe, Shi, Jie, Hu, Xiaolin
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Li, Xiao
Sun, Wenxuan
Chen, Huanran
Li, Qiongxiu
Liu, Yining
He, Yingzhe
Shi, Jie
Hu, Xiaolin
description Recently Diffusion-based Purification (DiffPure) has been recognized as an effective defense method against adversarial examples. However, we find DiffPure which directly employs the original pre-trained diffusion models for adversarial purification, to be suboptimal. This is due to an inherent trade-off between noise purification performance and data recovery quality. Additionally, the reliability of existing evaluations for DiffPure is questionable, as they rely on weak adaptive attacks. In this work, we propose a novel Adversarial Diffusion Bridge Model, termed ADBM. ADBM directly constructs a reverse bridge from the diffused adversarial data back to its original clean examples, enhancing the purification capabilities of the original diffusion models. Through theoretical analysis and experimental validation across various scenarios, ADBM has proven to be a superior and robust defense mechanism, offering significant promise for practical applications.
doi_str_mv 10.48550/arxiv.2408.00315
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2408_00315</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2408_00315</sourcerecordid><originalsourceid>FETCH-arxiv_primary_2408_003153</originalsourceid><addsrcrecordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjGw0DMwMDY05WTwc3Rx8rVScEwpSy0qTizKTMxRSMlMSystzszPU0gqykxJT1XIzU9JzVFIyy9SKErNyUxMyklVSERSX1BalJmWmZxYAtTCw8CalphTnMoLpbkZ5N1cQ5w9dME2xxcUZeYmFlXGg1wQD3aBMWEVAFq0O-o</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>ADBM: Adversarial diffusion bridge model for reliable adversarial purification</title><source>arXiv.org</source><creator>Li, Xiao ; Sun, Wenxuan ; Chen, Huanran ; Li, Qiongxiu ; Liu, Yining ; He, Yingzhe ; Shi, Jie ; Hu, Xiaolin</creator><creatorcontrib>Li, Xiao ; Sun, Wenxuan ; Chen, Huanran ; Li, Qiongxiu ; Liu, Yining ; He, Yingzhe ; Shi, Jie ; Hu, Xiaolin</creatorcontrib><description>Recently Diffusion-based Purification (DiffPure) has been recognized as an effective defense method against adversarial examples. However, we find DiffPure which directly employs the original pre-trained diffusion models for adversarial purification, to be suboptimal. This is due to an inherent trade-off between noise purification performance and data recovery quality. Additionally, the reliability of existing evaluations for DiffPure is questionable, as they rely on weak adaptive attacks. In this work, we propose a novel Adversarial Diffusion Bridge Model, termed ADBM. ADBM directly constructs a reverse bridge from the diffused adversarial data back to its original clean examples, enhancing the purification capabilities of the original diffusion models. Through theoretical analysis and experimental validation across various scenarios, ADBM has proven to be a superior and robust defense mechanism, offering significant promise for practical applications.</description><identifier>DOI: 10.48550/arxiv.2408.00315</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Learning</subject><creationdate>2024-08</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,782,887</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2408.00315$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2408.00315$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Li, Xiao</creatorcontrib><creatorcontrib>Sun, Wenxuan</creatorcontrib><creatorcontrib>Chen, Huanran</creatorcontrib><creatorcontrib>Li, Qiongxiu</creatorcontrib><creatorcontrib>Liu, Yining</creatorcontrib><creatorcontrib>He, Yingzhe</creatorcontrib><creatorcontrib>Shi, Jie</creatorcontrib><creatorcontrib>Hu, Xiaolin</creatorcontrib><title>ADBM: Adversarial diffusion bridge model for reliable adversarial purification</title><description>Recently Diffusion-based Purification (DiffPure) has been recognized as an effective defense method against adversarial examples. However, we find DiffPure which directly employs the original pre-trained diffusion models for adversarial purification, to be suboptimal. This is due to an inherent trade-off between noise purification performance and data recovery quality. Additionally, the reliability of existing evaluations for DiffPure is questionable, as they rely on weak adaptive attacks. In this work, we propose a novel Adversarial Diffusion Bridge Model, termed ADBM. ADBM directly constructs a reverse bridge from the diffused adversarial data back to its original clean examples, enhancing the purification capabilities of the original diffusion models. Through theoretical analysis and experimental validation across various scenarios, ADBM has proven to be a superior and robust defense mechanism, offering significant promise for practical applications.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Computer Vision and Pattern Recognition</subject><subject>Computer Science - Learning</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjGw0DMwMDY05WTwc3Rx8rVScEwpSy0qTizKTMxRSMlMSystzszPU0gqykxJT1XIzU9JzVFIyy9SKErNyUxMyklVSERSX1BalJmWmZxYAtTCw8CalphTnMoLpbkZ5N1cQ5w9dME2xxcUZeYmFlXGg1wQD3aBMWEVAFq0O-o</recordid><startdate>20240801</startdate><enddate>20240801</enddate><creator>Li, Xiao</creator><creator>Sun, Wenxuan</creator><creator>Chen, Huanran</creator><creator>Li, Qiongxiu</creator><creator>Liu, Yining</creator><creator>He, Yingzhe</creator><creator>Shi, Jie</creator><creator>Hu, Xiaolin</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20240801</creationdate><title>ADBM: Adversarial diffusion bridge model for reliable adversarial purification</title><author>Li, Xiao ; Sun, Wenxuan ; Chen, Huanran ; Li, Qiongxiu ; Liu, Yining ; He, Yingzhe ; Shi, Jie ; Hu, Xiaolin</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-arxiv_primary_2408_003153</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Computer Vision and Pattern Recognition</topic><topic>Computer Science - Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>Li, Xiao</creatorcontrib><creatorcontrib>Sun, Wenxuan</creatorcontrib><creatorcontrib>Chen, Huanran</creatorcontrib><creatorcontrib>Li, Qiongxiu</creatorcontrib><creatorcontrib>Liu, Yining</creatorcontrib><creatorcontrib>He, Yingzhe</creatorcontrib><creatorcontrib>Shi, Jie</creatorcontrib><creatorcontrib>Hu, Xiaolin</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Li, Xiao</au><au>Sun, Wenxuan</au><au>Chen, Huanran</au><au>Li, Qiongxiu</au><au>Liu, Yining</au><au>He, Yingzhe</au><au>Shi, Jie</au><au>Hu, Xiaolin</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>ADBM: Adversarial diffusion bridge model for reliable adversarial purification</atitle><date>2024-08-01</date><risdate>2024</risdate><abstract>Recently Diffusion-based Purification (DiffPure) has been recognized as an effective defense method against adversarial examples. However, we find DiffPure which directly employs the original pre-trained diffusion models for adversarial purification, to be suboptimal. This is due to an inherent trade-off between noise purification performance and data recovery quality. Additionally, the reliability of existing evaluations for DiffPure is questionable, as they rely on weak adaptive attacks. In this work, we propose a novel Adversarial Diffusion Bridge Model, termed ADBM. ADBM directly constructs a reverse bridge from the diffused adversarial data back to its original clean examples, enhancing the purification capabilities of the original diffusion models. Through theoretical analysis and experimental validation across various scenarios, ADBM has proven to be a superior and robust defense mechanism, offering significant promise for practical applications.</abstract><doi>10.48550/arxiv.2408.00315</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2408.00315
ispartof
issn
language eng
recordid cdi_arxiv_primary_2408_00315
source arXiv.org
subjects Computer Science - Artificial Intelligence
Computer Science - Computer Vision and Pattern Recognition
Computer Science - Learning
title ADBM: Adversarial diffusion bridge model for reliable adversarial purification
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-04T20%3A13%3A37IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=ADBM:%20Adversarial%20diffusion%20bridge%20model%20for%20reliable%20adversarial%20purification&rft.au=Li,%20Xiao&rft.date=2024-08-01&rft_id=info:doi/10.48550/arxiv.2408.00315&rft_dat=%3Carxiv_GOX%3E2408_00315%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true