{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,12]],"date-time":"2026-06-12T13:52:02Z","timestamp":1781272322711,"version":"3.54.1"},"reference-count":75,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","license":[{"start":{"date-parts":[[2025,1,1]],"date-time":"2025-01-01T00:00:00Z","timestamp":1735689600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,1,1]],"date-time":"2025-01-01T00:00:00Z","timestamp":1735689600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,1,1]],"date-time":"2025-01-01T00:00:00Z","timestamp":1735689600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62102092"],"award-info":[{"award-number":["62102092"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Multimedia"],"published-print":{"date-parts":[[2025]]},"DOI":"10.1109\/tmm.2024.3521721","type":"journal-article","created":{"date-parts":[[2024,12,23]],"date-time":"2024-12-23T19:23:06Z","timestamp":1734981786000},"page":"1374-1385","source":"Crossref","is-referenced-by-count":4,"title":["The Role of ViT Design and Training in Robustness to Common Corruptions"],"prefix":"10.1109","volume":"27","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9074-0667","authenticated-orcid":false,"given":"Rui","family":"Tian","sequence":"first","affiliation":[{"name":"Shanghai Key Laboratory of Intelligent Information Processing, School of Computer Science, Fudan University, Shanghai, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8689-5807","authenticated-orcid":false,"given":"Zuxuan","family":"Wu","sequence":"additional","affiliation":[{"name":"Shanghai Key Laboratory of Intelligent Information Processing, School of Computer Science, Fudan University, Shanghai, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4693-2968","authenticated-orcid":false,"given":"Qi","family":"Dai","sequence":"additional","affiliation":[{"name":"Microsoft Research Asia, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Micah","family":"Goldblum","sequence":"additional","affiliation":[{"name":"Center for Data Science, New York University, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5104-6146","authenticated-orcid":false,"given":"Han","family":"Hu","sequence":"additional","affiliation":[{"name":"Microsoft Research Asia, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1907-8567","authenticated-orcid":false,"given":"Yu-Gang","family":"Jiang","sequence":"additional","affiliation":[{"name":"Shanghai Key Laboratory of Intelligent Information Processing, School of Computer Science, Fudan University, Shanghai, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","first-page":"26831","article-title":"Are transformers more robust than CNNs?","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Bai","year":"2021"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-20053-3_1"},{"key":"ref3","article-title":"MultiGrain: A unified image embedding for classes and instances","author":"Berman","year":"2019"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.01007"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00051"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01104"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.195"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-020-01383-2"},{"key":"ref9","article-title":"Conditional positional encodings for vision transformers","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Chu","year":"2023"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW50498.2020.00359"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr.2019.00020"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3278694"},{"key":"ref13","first-page":"7480","article-title":"Scaling vision transformers to 22 billion parameters","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Dehghani","year":"2023"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00105"},{"key":"ref15","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Dosovitskiy","year":"2021"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33013714"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00400"},{"key":"ref18","first-page":"2633","article-title":"Global context vision transformers","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hatamizadeh","year":"2023"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref20","article-title":"Benchmarking neural network robustness to common corruptions and perturbations","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hendrycks","year":"2018"},{"key":"ref21","article-title":"AugMix: A simple data processing method to improve robustness and uncertainty","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hendrycks","year":"2019"},{"key":"ref22","first-page":"15262","article-title":"Natural adversarial examples","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit.","author":"Hendrycks","year":"2021"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.01172"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01628"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00823"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01306"},{"key":"ref27","article-title":"How much position information do convolutional neural networks encode?","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Islam","year":"2019"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/TMM.2022.3148592"},{"key":"ref29","first-page":"18941","article-title":"3D common corruptions and data augmentation","volume-title":"Proc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit.","author":"Fatih Kar","year":"2022"},{"key":"ref30","article-title":"Adam: A method for stochastic optimization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Kingma","year":"2015"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3065386"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01220"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2022.3164083"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00986"},{"key":"ref35","article-title":"Decoupled weight decay regularization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Loshchilov","year":"2018"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00774"},{"key":"ref37","article-title":"Discrete representations strengthen vision transformer robustness","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Mao","year":"2022"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01173"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19806-9_36"},{"key":"ref40","first-page":"23296","article-title":"Intriguing properties of vision transformers","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Naseer","year":"2021"},{"key":"ref41","article-title":"Robustmix: Improving robustness by regularizing the frequency bias of deep nets","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst. Workshop","author":"Ngnawe","year":"2022"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2105.07581"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19778-9_27"},{"key":"ref44","first-page":"13937","article-title":"DynamicViT: Efficient vision transformers with dynamic token sparsification","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Rao","year":"2021"},{"key":"ref45","article-title":"PyTorch image models","volume-title":"GitHub repository","author":"Wightman","year":"2019"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58580-8_4"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-015-0816-y"},{"key":"ref48","first-page":"3353","article-title":"Adversarial training for free!","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Shafahi","year":"2019"},{"key":"ref49","article-title":"On the adversarial robustness of vision transformers","volume-title":"Trans. Mach. Learn. Res.","author":"Shao","year":"2022"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00424"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/ICRA48506.2021.9561205"},{"key":"ref52","first-page":"28042","article-title":"Encoding robustness to image style via adversarial feature perturbations","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Shu","year":"2021"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01063"},{"key":"ref54","first-page":"10347","article-title":"Training data-efficient image transformers & distillation through attention","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Touvron","year":"2021"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-20053-3_30"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00723"},{"key":"ref57","first-page":"10506","article-title":"Learning robust global representations by penalizing local predictive power","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Wang","year":"2019"},{"key":"ref58","first-page":"237","article-title":"AugMax: Adversarial composition of random augmentations for robust training","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Wang","year":"2021"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00061"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1007\/s41095-022-0274-8"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00009"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00090"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/TMM.2022.3232030"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3268446"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3379457"},{"key":"ref66","first-page":"1305","article-title":"CondConv: Conditionally parameterized convolutions for efficient inference","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Yang","year":"2019"},{"key":"ref67","first-page":"13255","article-title":"A Fourier perspective on model robustness in computer vision","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Yin","year":"2019"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00062"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00612"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01179"},{"key":"ref71","article-title":"MixUp: Beyond empirical risk minimization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Zhang","year":"2018"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/TMM.2021.3119994"},{"key":"ref73","article-title":"DyNet: Dynamic convolution for accelerating convolutional neural networks","author":"Zhang","year":"2020"},{"key":"ref74","first-page":"27378","article-title":"Understanding the robustness in vision transformers","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zhou","year":"2022"},{"key":"ref75","article-title":"Dynamic resolution network","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Zhu","year":"2021"}],"container-title":["IEEE Transactions on Multimedia"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/6046\/10844992\/10812859.pdf?arnumber=10812859","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,11]],"date-time":"2025-03-11T17:44:43Z","timestamp":1741715083000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10812859\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025]]},"references-count":75,"URL":"https:\/\/doi.org\/10.1109\/tmm.2024.3521721","relation":{},"ISSN":["1520-9210","1941-0077"],"issn-type":[{"value":"1520-9210","type":"print"},{"value":"1941-0077","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025]]}}}