Unsupervised Anomaly Detection via Masked Diffusion Posterior Sampling
Reconstruction-based methods have been commonly used for unsupervised anomaly detection, in which a normal image is reconstructed and compared with the given test image to detect and locate anomalies. Recently, diffusion models have shown promising applications for anomaly detection due to their pow...
Gespeichert in:
Veröffentlicht in: | arXiv.org 2024-04 |
---|---|
Hauptverfasser: | , , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | |
container_start_page | |
container_title | arXiv.org |
container_volume | |
creator | Wu, Di Fan, Shicai Zhou, Xue Li, Yu Deng, Yuzhong Zou, Jianxiao Lin, Baihong |
description | Reconstruction-based methods have been commonly used for unsupervised anomaly detection, in which a normal image is reconstructed and compared with the given test image to detect and locate anomalies. Recently, diffusion models have shown promising applications for anomaly detection due to their powerful generative ability. However, these models lack strict mathematical support for normal image reconstruction and unexpectedly suffer from low reconstruction quality. To address these issues, this paper proposes a novel and highly-interpretable method named Masked Diffusion Posterior Sampling (MDPS). In MDPS, the problem of normal image reconstruction is mathematically modeled as multiple diffusion posterior sampling for normal images based on the devised masked noisy observation model and the diffusion-based normal image prior under Bayesian framework. Using a metric designed from pixel-level and perceptual-level perspectives, MDPS can effectively compute the difference map between each normal posterior sample and the given test image. Anomaly scores are obtained by averaging all difference maps for multiple posterior samples. Exhaustive experiments on MVTec and BTAD datasets demonstrate that MDPS can achieve state-of-the-art performance in normal image reconstruction quality as well as anomaly detection and localization. |
format | Article |
fullrecord | <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_3049779032</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>3049779032</sourcerecordid><originalsourceid>FETCH-proquest_journals_30497790323</originalsourceid><addsrcrecordid>eNqNitEKgjAUQEcQJOU_DHoW1qaZj5FJL0FQPcuoa8x0W7ub0N9n0Af0dOCcMyERF2KVbFLOZyRGbBljfJ3zLBMRqa4agwU3KIQ73WrTy-5NS_Bw88poOihJjxKfYyxV0wT8ypNBD04ZR8-yt53SjwWZNrJDiH-ck2W1v-wOiXXmFQB93Zrg9JhqwdIizwsmuPjv-gD7Iju-</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>3049779032</pqid></control><display><type>article</type><title>Unsupervised Anomaly Detection via Masked Diffusion Posterior Sampling</title><source>Free E- Journals</source><creator>Wu, Di ; Fan, Shicai ; Zhou, Xue ; Li, Yu ; Deng, Yuzhong ; Zou, Jianxiao ; Lin, Baihong</creator><creatorcontrib>Wu, Di ; Fan, Shicai ; Zhou, Xue ; Li, Yu ; Deng, Yuzhong ; Zou, Jianxiao ; Lin, Baihong</creatorcontrib><description>Reconstruction-based methods have been commonly used for unsupervised anomaly detection, in which a normal image is reconstructed and compared with the given test image to detect and locate anomalies. Recently, diffusion models have shown promising applications for anomaly detection due to their powerful generative ability. However, these models lack strict mathematical support for normal image reconstruction and unexpectedly suffer from low reconstruction quality. To address these issues, this paper proposes a novel and highly-interpretable method named Masked Diffusion Posterior Sampling (MDPS). In MDPS, the problem of normal image reconstruction is mathematically modeled as multiple diffusion posterior sampling for normal images based on the devised masked noisy observation model and the diffusion-based normal image prior under Bayesian framework. Using a metric designed from pixel-level and perceptual-level perspectives, MDPS can effectively compute the difference map between each normal posterior sample and the given test image. Anomaly scores are obtained by averaging all difference maps for multiple posterior samples. Exhaustive experiments on MVTec and BTAD datasets demonstrate that MDPS can achieve state-of-the-art performance in normal image reconstruction quality as well as anomaly detection and localization.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Anomalies ; Image quality ; Image reconstruction ; Sampling</subject><ispartof>arXiv.org, 2024-04</ispartof><rights>2024. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</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>776,780</link.rule.ids></links><search><creatorcontrib>Wu, Di</creatorcontrib><creatorcontrib>Fan, Shicai</creatorcontrib><creatorcontrib>Zhou, Xue</creatorcontrib><creatorcontrib>Li, Yu</creatorcontrib><creatorcontrib>Deng, Yuzhong</creatorcontrib><creatorcontrib>Zou, Jianxiao</creatorcontrib><creatorcontrib>Lin, Baihong</creatorcontrib><title>Unsupervised Anomaly Detection via Masked Diffusion Posterior Sampling</title><title>arXiv.org</title><description>Reconstruction-based methods have been commonly used for unsupervised anomaly detection, in which a normal image is reconstructed and compared with the given test image to detect and locate anomalies. Recently, diffusion models have shown promising applications for anomaly detection due to their powerful generative ability. However, these models lack strict mathematical support for normal image reconstruction and unexpectedly suffer from low reconstruction quality. To address these issues, this paper proposes a novel and highly-interpretable method named Masked Diffusion Posterior Sampling (MDPS). In MDPS, the problem of normal image reconstruction is mathematically modeled as multiple diffusion posterior sampling for normal images based on the devised masked noisy observation model and the diffusion-based normal image prior under Bayesian framework. Using a metric designed from pixel-level and perceptual-level perspectives, MDPS can effectively compute the difference map between each normal posterior sample and the given test image. Anomaly scores are obtained by averaging all difference maps for multiple posterior samples. Exhaustive experiments on MVTec and BTAD datasets demonstrate that MDPS can achieve state-of-the-art performance in normal image reconstruction quality as well as anomaly detection and localization.</description><subject>Anomalies</subject><subject>Image quality</subject><subject>Image reconstruction</subject><subject>Sampling</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>BENPR</sourceid><recordid>eNqNitEKgjAUQEcQJOU_DHoW1qaZj5FJL0FQPcuoa8x0W7ub0N9n0Af0dOCcMyERF2KVbFLOZyRGbBljfJ3zLBMRqa4agwU3KIQ73WrTy-5NS_Bw88poOihJjxKfYyxV0wT8ypNBD04ZR8-yt53SjwWZNrJDiH-ck2W1v-wOiXXmFQB93Zrg9JhqwdIizwsmuPjv-gD7Iju-</recordid><startdate>20240427</startdate><enddate>20240427</enddate><creator>Wu, Di</creator><creator>Fan, Shicai</creator><creator>Zhou, Xue</creator><creator>Li, Yu</creator><creator>Deng, Yuzhong</creator><creator>Zou, Jianxiao</creator><creator>Lin, Baihong</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20240427</creationdate><title>Unsupervised Anomaly Detection via Masked Diffusion Posterior Sampling</title><author>Wu, Di ; Fan, Shicai ; Zhou, Xue ; Li, Yu ; Deng, Yuzhong ; Zou, Jianxiao ; Lin, Baihong</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_30497790323</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Anomalies</topic><topic>Image quality</topic><topic>Image reconstruction</topic><topic>Sampling</topic><toplevel>online_resources</toplevel><creatorcontrib>Wu, Di</creatorcontrib><creatorcontrib>Fan, Shicai</creatorcontrib><creatorcontrib>Zhou, Xue</creatorcontrib><creatorcontrib>Li, Yu</creatorcontrib><creatorcontrib>Deng, Yuzhong</creatorcontrib><creatorcontrib>Zou, Jianxiao</creatorcontrib><creatorcontrib>Lin, Baihong</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science & Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Wu, Di</au><au>Fan, Shicai</au><au>Zhou, Xue</au><au>Li, Yu</au><au>Deng, Yuzhong</au><au>Zou, Jianxiao</au><au>Lin, Baihong</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Unsupervised Anomaly Detection via Masked Diffusion Posterior Sampling</atitle><jtitle>arXiv.org</jtitle><date>2024-04-27</date><risdate>2024</risdate><eissn>2331-8422</eissn><abstract>Reconstruction-based methods have been commonly used for unsupervised anomaly detection, in which a normal image is reconstructed and compared with the given test image to detect and locate anomalies. Recently, diffusion models have shown promising applications for anomaly detection due to their powerful generative ability. However, these models lack strict mathematical support for normal image reconstruction and unexpectedly suffer from low reconstruction quality. To address these issues, this paper proposes a novel and highly-interpretable method named Masked Diffusion Posterior Sampling (MDPS). In MDPS, the problem of normal image reconstruction is mathematically modeled as multiple diffusion posterior sampling for normal images based on the devised masked noisy observation model and the diffusion-based normal image prior under Bayesian framework. Using a metric designed from pixel-level and perceptual-level perspectives, MDPS can effectively compute the difference map between each normal posterior sample and the given test image. Anomaly scores are obtained by averaging all difference maps for multiple posterior samples. Exhaustive experiments on MVTec and BTAD datasets demonstrate that MDPS can achieve state-of-the-art performance in normal image reconstruction quality as well as anomaly detection and localization.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | EISSN: 2331-8422 |
ispartof | arXiv.org, 2024-04 |
issn | 2331-8422 |
language | eng |
recordid | cdi_proquest_journals_3049779032 |
source | Free E- Journals |
subjects | Anomalies Image quality Image reconstruction Sampling |
title | Unsupervised Anomaly Detection via Masked Diffusion Posterior Sampling |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-27T05%3A27%3A52IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Unsupervised%20Anomaly%20Detection%20via%20Masked%20Diffusion%20Posterior%20Sampling&rft.jtitle=arXiv.org&rft.au=Wu,%20Di&rft.date=2024-04-27&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E3049779032%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=3049779032&rft_id=info:pmid/&rfr_iscdi=true |