{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,10]],"date-time":"2026-02-10T16:03:07Z","timestamp":1770739387939,"version":"3.49.0"},"reference-count":84,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"3","license":[{"start":{"date-parts":[[2026,3,1]],"date-time":"2026-03-01T00:00:00Z","timestamp":1772323200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,3,1]],"date-time":"2026-03-01T00:00:00Z","timestamp":1772323200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,3,1]],"date-time":"2026-03-01T00:00:00Z","timestamp":1772323200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100021171","name":"Basic and Applied Basic Research Foundation of Guangdong Province","doi-asserted-by":"publisher","award":["2024B1515020095"],"award-info":[{"award-number":["2024B1515020095"]}],"id":[{"id":"10.13039\/501100021171","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Guangdong Provincial Program","award":["2023TQ07A352"],"award-info":[{"award-number":["2023TQ07A352"]}]},{"name":"Sub-topic of Key R&D Projects of the Ministry of Science and Technology","award":["2023YFC3304804"],"award-info":[{"award-number":["2023YFC3304804"]}]},{"name":"Shenzhen Science and Technology Program","award":["RCYX20210609103057050"],"award-info":[{"award-number":["RCYX20210609103057050"]}]},{"name":"Shenzhen Science and Technology Program","award":["JCYJ20240813113608011"],"award-info":[{"award-number":["JCYJ20240813113608011"]}]},{"name":"Longgang District Key Laboratory of Intelligent Digital Economy Security"},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62471420"],"award-info":[{"award-number":["62471420"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100021171","name":"Basic and Applied Basic Research Foundation of Guangdong Province","doi-asserted-by":"publisher","award":["2025A1515012296"],"award-info":[{"award-number":["2025A1515012296"]}],"id":[{"id":"10.13039\/501100021171","id-type":"DOI","asserted-by":"publisher"}]},{"name":"2025 Tencent AI Lab Rhino-Bird Program"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2026,3]]},"DOI":"10.1109\/tpami.2025.3642161","type":"journal-article","created":{"date-parts":[[2025,12,9]],"date-time":"2025-12-09T18:32:53Z","timestamp":1765305173000},"page":"3680-3697","source":"Crossref","is-referenced-by-count":0,"title":["Versatile Backdoor Attack With Visible, Semantic, Sample-Specific and Compatible Triggers"],"prefix":"10.1109","volume":"48","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-2761-7997","authenticated-orcid":false,"given":"Ruotong","family":"Wang","sequence":"first","affiliation":[{"name":"Chinese University of Hong Kong, Shenzhen, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-5997-3041","authenticated-orcid":false,"given":"Hongrui","family":"Chen","sequence":"additional","affiliation":[{"name":"Chinese University of Hong Kong, Shenzhen, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1225-1718","authenticated-orcid":false,"given":"Zihao","family":"Zhu","sequence":"additional","affiliation":[{"name":"Chinese University of Hong Kong, Shenzhen, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4497-0135","authenticated-orcid":false,"given":"Li","family":"Liu","sequence":"additional","affiliation":[{"name":"Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2183-5990","authenticated-orcid":false,"given":"Baoyuan","family":"Wu","sequence":"additional","affiliation":[{"name":"Chinese University of Hong Kong, Shenzhen, China"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2014.2343217"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3345880"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3052549"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2022.3200245"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2018.2889052"},{"key":"ref6","article-title":"Attacks in adversarial machine learning: A systematic survey from the life-cycle perspective","author":"Wu","year":"2025","journal-title":"Int. J. Comput. Vis."},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/TIFS.2024.3402179"},{"key":"ref8","first-page":"28289","article-title":"BadChain: Backdoor chain-of-thought prompting for large language models","volume-title":"Proc. 12th Int. Conf. Learn. Representations","author":"Xiang","year":"2024"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-025-02368-9"},{"key":"ref10","first-page":"19075","article-title":"BadVideo: Stealthy backdoor attack against text-to-video generation","volume-title":"Proc. Int. Conf. Comput. Vis.","author":"Wang","year":"2025"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2019.2909068"},{"key":"ref12","article-title":"Targeted backdoor attacks on deep learning systems using data poisoning","author":"Chen","year":"2017"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01465"},{"key":"ref14","first-page":"3454","article-title":"Input-aware dynamic backdoor attack","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Nguyen","year":"2020"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.14722\/ndss.2018.23291"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ICIP.2019.8802997"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.01615"},{"key":"ref18","first-page":"1","article-title":"WaNet\u2013imperceptible warping-based backdoor attack","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Nguyen","year":"2021"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/TIFS.2023.3333687"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00614"},{"key":"ref21","article-title":"Dangerous cloaking: Natural trigger based backdoor attacks on object detectors in the physical world","author":"Ma","year":"2022"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/SRDS60354.2023.00018"},{"key":"ref23","article-title":"Label-consistent backdoor attacks","author":"Turner","year":"2019"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.01616"},{"key":"ref25","first-page":"1505","article-title":"Blind backdoors in deep learning models","volume-title":"Proc. USENIX Secur. Symp.","author":"Bagdasaryan","year":"2021"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-25056-9_26"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1145\/3560830.3563730"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49357.2023.10095980"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2023.3342922"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1186\/s13635-020-00104-z"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3372297.3423362"},{"key":"ref32","article-title":"Facehack: Triggering backdoored facial recognition systems using facial characteristics","author":"Sarkar","year":"2020"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1007\/s12083-020-01031-z"},{"key":"ref34","article-title":"Light can hack your face! Black-box backdoor attack on face recognition systems","author":"Li","year":"2020"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2025.3611340"},{"key":"ref36","first-page":"8011","article-title":"Spectral signatures in backdoor attacks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Tran","year":"2018"},{"key":"ref37","first-page":"18670","article-title":"VDC: Versatile data cleanser based on visual-linguistic inconsistency by multimodal large language models","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Zhu","year":"2024"},{"key":"ref38","first-page":"1685","article-title":"Towards a proactive $\\lbrace${ML$\\rbrace$} approach for detecting backdoor poison samples","volume-title":"Proc. USENIX Secur. Symp.","author":"Qi","year":"2023"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1145\/3427228.3427264"},{"key":"ref40","first-page":"76161","article-title":"Activation gradient based poisoned sample detection against backdoor attacks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Yuan","year":"2025"},{"key":"ref41","article-title":"Reliable poisoned sample detection against backdoor attacks enhanced by sharpness aware minimization","author":"Zhang","year":"2024"},{"key":"ref42","first-page":"14900","article-title":"Anti-backdoor learning: Training clean models on poisoned data","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Li","year":"2021"},{"key":"ref43","first-page":"1","article-title":"Backdoor defense via decoupling the training process","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Huang","year":"2022"},{"key":"ref44","first-page":"9727","article-title":"Effective backdoor defense by exploiting sensitivity of poisoned samples","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"35","author":"Chen","year":"2022"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.52202\/079017-2565"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-00470-5_13"},{"key":"ref47","article-title":"Detecting backdoor attacks on deep neural networks by activation clustering","volume-title":"Proc. AAAI Workshop Artif. Intell. Saf.","author":"Chen","year":"2019"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-20065-6_11"},{"key":"ref49","first-page":"1","article-title":"Neural attention distillation: Erasing backdoor triggers from deep neural networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Li","year":"2021"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/SP.2019.00031"},{"key":"ref51","first-page":"1132","article-title":"Neural polarizer: A lightweight and effective backdoor defense via purifying poisoned features","volume-title":"Proc. 37th Conf. Neural Inf. Process. Syst.","author":"Zhu","year":"2023"},{"key":"ref52","first-page":"25876","article-title":"Shared adversarial unlearning: Backdoor mitigation by unlearning shared adversarial examples","volume-title":"Proc. Conf. Neural Inf. Process. Syst.","author":"Wei","year":"2023"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1332"},{"key":"ref54","first-page":"18667","article-title":"Pre-activation distributions expose backdoor neurons","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"35","author":"Zheng","year":"2022"},{"key":"ref55","first-page":"2938","article-title":"How to backdoor federated learning","volume-title":"Proc. Int. Conf. Artif. Intell. Statist.","author":"Bagdasaryan","year":"2020"},{"key":"ref56","first-page":"22103","article-title":"Finding naturally occurring physical backdoors in image datasets","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Wenger","year":"2022"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1145\/3503161.3548171"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.74"},{"key":"ref59","article-title":"Rethinking the trigger of backdoor attack","author":"Li","year":"2020"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1109\/TDSC.2020.3021407"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00786"},{"key":"ref62","article-title":"Gpt-4 technical report","author":"Achiam","year":"2023"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00585"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01208"},{"key":"ref65","article-title":"IP-adapter: Text compatible image prompt adapter for text-to-image diffusion models","author":"Ye","year":"2023"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00355"},{"key":"ref67","first-page":"34892","article-title":"Visual instruction tuning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Liu","year":"2024"},{"key":"ref68","article-title":"Backdoor attack in the physical world","author":"Li","year":"2021"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v35i10.17037"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1145\/2986035.2986039"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref73","first-page":"1","article-title":"Very deep convolutional networks for large-scale image recognition","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Simonyan","year":"2015"},{"key":"ref74","first-page":"10546","article-title":"BackdoorBench: A comprehensive benchmark of backdoor learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"35","author":"Wu","year":"2022"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-025-02447-x"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-009-0275-4"},{"key":"ref77","article-title":"YOLOv4: Optimal speed and accuracy of object detection","author":"Bochkovskiy","year":"2020"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2016.2577031"},{"key":"ref79","article-title":"Learning face representation from scratch","author":"Yi","year":"2014"},{"key":"ref80","article-title":"Labeled faces in the wild: A database forstudying face recognition in unconstrained environments","author":"Huang","year":"2007"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01764"},{"key":"ref82","first-page":"16913","article-title":"Adversarial neuron pruning purifies backdoored deep models","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Wu","year":"2021"},{"key":"ref83","first-page":"18667","article-title":"Pre-activation distributions expose backdoor neurons","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Zheng","year":"2022"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1145\/3359789.3359790"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/34\/11372200\/11291169.pdf?arnumber=11291169","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,9]],"date-time":"2026-02-09T21:06:04Z","timestamp":1770671164000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11291169\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3]]},"references-count":84,"journal-issue":{"issue":"3"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2025.3642161","relation":{},"ISSN":["0162-8828","2160-9292","1939-3539"],"issn-type":[{"value":"0162-8828","type":"print"},{"value":"2160-9292","type":"electronic"},{"value":"1939-3539","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3]]}}}