{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,25]],"date-time":"2026-07-25T02:18:59Z","timestamp":1784945939544,"version":"3.55.0"},"reference-count":105,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"8","license":[{"start":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T00:00:00Z","timestamp":1785542400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T00:00:00Z","timestamp":1785542400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T00:00:00Z","timestamp":1785542400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"New Generation Artificial Intelligence-National Science and Technology Major Project","award":["2025ZD0123502"],"award-info":[{"award-number":["2025ZD0123502"]}]},{"name":"Shenzhen Science and Technology Program","award":["JCYJ20220818103006012"],"award-info":[{"award-number":["JCYJ20220818103006012"]}]},{"name":"Shenzhen Science and Technology Program","award":["KJZD20240903100202004"],"award-info":[{"award-number":["KJZD20240903100202004"]}]},{"DOI":"10.13039\/501100021171","name":"Guangdong Basic and Applied Basic Research Foundation","doi-asserted-by":"crossref","award":["2024B1515020095"],"award-info":[{"award-number":["2024B1515020095"]}],"id":[{"id":"10.13039\/501100021171","id-type":"DOI","asserted-by":"crossref"}]},{"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"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2026,8]]},"DOI":"10.1109\/tpami.2026.3679726","type":"journal-article","created":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T20:12:30Z","timestamp":1775074350000},"page":"9552-9570","source":"Crossref","is-referenced-by-count":2,"title":["Seeking Flat Minima Over Diverse Surrogates for Improved Adversarial Transferability: A Theoretical Framework and Algorithmic Instantiation"],"prefix":"10.1109","volume":"48","author":[{"ORCID":"https:\/\/orcid.org\/0009-0008-9955-9117","authenticated-orcid":false,"given":"Meixi","family":"Zheng","sequence":"first","affiliation":[{"name":"School of Artificial Intelligence, Chinese University of Hong Kong, Shenzhen, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kehan","family":"Wu","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence, Chinese University of Hong Kong, Shenzhen, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8530-485X","authenticated-orcid":false,"given":"Yanbo","family":"Fan","sequence":"additional","affiliation":[{"name":"Nanjing University (Suzhou Campus), Suzhou, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7950-1662","authenticated-orcid":false,"given":"Rui","family":"Huang","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence, Chinese University of Hong Kong, Shenzhen, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2183-5990","authenticated-orcid":false,"given":"Baoyuan","family":"Wu","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence, Chinese University of Hong Kong, Shenzhen, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01467"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2022.3194988"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2022.3169802"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3126733"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-025-02692-0"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3480519"},{"key":"ref7","first-page":"2854","article-title":"Nesterov accelerated gradient and scale invariance for adversarial attacks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lin"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00284"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00957"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00196"},{"key":"ref11","first-page":"171","article-title":"Rethinking model ensemble in transfer-based adversarial attacks","volume-title":"Proc. 12th Int. Conf. Learn. Representations","author":"Chen"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01965"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/TMM.2023.3315550"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.2022.3160659"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2164"},{"key":"ref16","article-title":"Transferability in machine learning: From phenomena to black-box attacks using adversarial samples","author":"Papernot","year":"2016"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1201\/9781351251389-8"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.5244\/C.35.186"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.52202\/075280-3073"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1325"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/tifs.2026.3658188"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/ICME55011.2023.00212"},{"key":"ref23","first-page":"2315","article-title":"Sharpness-aware minimization for efficiently improving generalization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Foret"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00483"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00754"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01457"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.01585"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00444"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19772-7_32"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00425"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2022.3211736"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i07.6810"},{"key":"ref33","first-page":"1","article-title":"Making substitute models more Bayesian can enhance transferability of adversarial examples","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Li"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00414"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02301"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01456"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/307400.307435"},{"key":"ref38","article-title":"The intriguing role of module criticality in the generalization of deep networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Chatterji"},{"key":"ref39","first-page":"191","article-title":"On information-type measure of difference of probability distributions and indirect observations","volume":"2","author":"Csisz\u00e1r","year":"1967","journal-title":"Studia Sci. Math. Hungar."},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.2016.2603151"},{"issue":"128","key":"ref41","first-page":"1","article-title":"Optimal bounds between F-divergences and integral probability metrics","volume":"22","author":"Agrawal","year":"2021","journal-title":"J. Mach. Learn. Res."},{"key":"ref42","article-title":"On localized discrepancy for domain adaptation","author":"Zhang","year":"2020"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0215"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.06083"},{"key":"ref45","first-page":"2874","article-title":"On large-batch training for deep learning: Generalization gap and sharp minima","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Keskar"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19772-7_35"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2025.3574432"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00774"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01055"},{"key":"ref50","first-page":"8803","article-title":"Loss surfaces, mode connectivity, and fast ensembling of DNNs","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Garipov"},{"key":"ref51","first-page":"4873","article-title":"Stochastic gradient descent as approximate bayesian inference","volume":"18","author":"Stephan","year":"2017","journal-title":"J. Mach. Learn. Res."},{"key":"ref52","first-page":"III-1139","article-title":"On the importance of initialization and momentum in deep learning","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Sutskever"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2025.3610085"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-015-0816-y"},{"key":"ref55","article-title":"Learning multiple layers of features from tiny images","author":"Krizhevsky","year":"2009"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1145\/3065386"},{"key":"ref57","article-title":"Very deep convolutional networks for large-scale image recognition","volume-title":"Proc. 3rd Int. Conf. Learn. Representations","author":"Simonyan"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.243"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.4324\/9781410605337-29"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01264-9_8"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00474"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00140"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00293"},{"key":"ref64","first-page":"6105","article-title":"Efficientnet: Rethinking model scaling for convolutional neural networks","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Tan"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01167"},{"key":"ref66","first-page":"611","article-title":"An image is worth 16\u2009x16 words: Transformers for image recognition at scale","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Dosovitskiy"},{"key":"ref67","first-page":"10347","article-title":"Training data-efficient image transformers & distillation through attention","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Touvron"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00060"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00986"},{"key":"ref70","first-page":"20014","article-title":"XCiT: Cross-covariance image transformers","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Ali"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00010"},{"key":"ref72","article-title":"Robust principles: Architectural design principles for adversarially robust CNNs","author":"Peng","year":"2023"},{"key":"ref73","first-page":"3533","article-title":"Do adversarially robust imagenet models transfer better?","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Salman"},{"key":"ref74","first-page":"5849","article-title":"Fast is better than free: Revisiting adversarial training","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Wong"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-024-02196-3"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0614"},{"key":"ref77","first-page":"1894","article-title":"Ensemble adversarial training: Attacks and defenses","volume-title":"Proc. 6th Int. Conf. Learn. Representations","author":"Tram\u00e8r"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1109\/SaTML54575.2023.00024"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref80","first-page":"26831","article-title":"Are transformers more robust than CNNs?","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Bai"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.2307\/1428011"},{"key":"ref82","doi-asserted-by":"publisher","DOI":"10.1214\/aos\/1015957395"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.634"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.5244\/C.30.87"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46493-0_38"},{"key":"ref86","first-page":"16048","article-title":"Understanding and improving fast adversarial training","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Andriushchenko"},{"key":"ref87","first-page":"11192","article-title":"Unlabeled data improves adversarial robustness","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Carmon"},{"key":"ref88","first-page":"19655","article-title":"HYDRA: Pruning adversarially robust neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Sehwag"},{"key":"ref89","article-title":"Improving adversarial robustness requires revisiting misclassified examples","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Wang"},{"key":"ref90","first-page":"8093","article-title":"Overfitting in adversarially robust deep learning","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Rice"},{"key":"ref91","first-page":"7472","article-title":"Theoretically principled trade-off between robustness and accuracy","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zhang"},{"key":"ref92","article-title":"Adversarial robustness as a prior for learned representations","author":"Engstrom","year":"2019"},{"key":"ref93","first-page":"19365","article-title":"Self-adaptive training: Beyond empirical risk minimization","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Huang"},{"key":"ref94","first-page":"36246","article-title":"Better diffusion models further improve adversarial training","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Wang"},{"key":"ref95","first-page":"41429","article-title":"Detecting adversarial data by probing multiple perturbations using expected perturbation score","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zhang"},{"key":"ref96","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1807"},{"key":"ref97","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2554"},{"key":"ref98","first-page":"8179","article-title":"Transferable adversarial attack based on integrated gradients","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Huang"},{"key":"ref99","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v38i6.28430"},{"key":"ref100","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3381180"},{"key":"ref101","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-024-02103-w"},{"key":"ref102","first-page":"1635","article-title":"A new defense against adversarial images: Turning a weakness into a strength","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Hu"},{"key":"ref103","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00758"},{"key":"ref104","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2025.3611340"},{"key":"ref105","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2440"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/34\/11595778\/11460256.pdf?arnumber=11460256","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T19:45:04Z","timestamp":1783453504000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11460256\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8]]},"references-count":105,"journal-issue":{"issue":"8"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2026.3679726","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":[[2026,8]]}}}