{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T19:52:57Z","timestamp":1791489177255,"version":"4.3.4"},"reference-count":88,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"11","license":[{"start":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T00:00:00Z","timestamp":1793491200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T00:00:00Z","timestamp":1793491200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T00:00:00Z","timestamp":1793491200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"Joint WeBank-NTU Research Institute on Fintech, Nanyang Technological University, Singapore","id":[{"id":"https:\/\/ror.org\/02e7b5302","id-type":"ROR","asserted-by":"crossref"}]},{"name":"National Natural Science Foundation of China","award":["U24A20327"],"award-info":[{"award-number":["U24A20327"]}],"id":[{"id":"https:\/\/ror.org\/01h0zpd94","id-type":"ROR","asserted-by":"crossref"}]},{"name":"Guangdong S&T Program","award":["2026B0101110001"],"award-info":[{"award-number":["2026B0101110001"]}]},{"name":"GuangDong Basic and Applied Basic Research Foundation","award":["2026A1515010388"],"award-info":[{"award-number":["2026A1515010388"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2026,11]]},"DOI":"10.1109\/tpami.2026.3721895","type":"journal-article","created":{"date-parts":[[2026,8,11]],"date-time":"2026-08-11T19:04:30Z","timestamp":1786475070000},"page":"14878-14895","source":"Crossref","is-referenced-by-count":1,"title":["Adapt in the Wild: Test-Time Entropy Minimization with Sharpness and Feature Regularization"],"prefix":"10.1109","volume":"48","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8212-1831","authenticated-orcid":false,"given":"Shuaicheng","family":"Niu","sequence":"first","affiliation":[{"name":"South China University of Technology","place":["Guangzhou, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Guohao","family":"Chen","sequence":"additional","affiliation":[{"name":"Nanyang Technological University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Deyu","family":"Chen","sequence":"additional","affiliation":[{"name":"South China University of Technology","place":["Guangzhou, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2125-1074","authenticated-orcid":false,"given":"Yifan","family":"Zhang","sequence":"additional","affiliation":[{"name":"National University of Singapore"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9132-5625","authenticated-orcid":false,"given":"Jiaxiang","family":"Wu","sequence":"additional","affiliation":[{"name":"ByteDance","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4969-0060","authenticated-orcid":false,"given":"Zhiquan","family":"Wen","sequence":"additional","affiliation":[{"name":"South China University of Technology","place":["Guangzhou, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8916-8368","authenticated-orcid":false,"given":"Yaofo","family":"Chen","sequence":"additional","affiliation":[{"name":"South China University of Technology","place":["Guangzhou, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8543-3953","authenticated-orcid":false,"given":"Peilin","family":"Zhao","sequence":"additional","affiliation":[{"name":"Shanghai Jiao Tong University","place":["Shanghai, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0300-3448","authenticated-orcid":false,"given":"Chunyan","family":"Miao","sequence":"additional","affiliation":[{"name":"Nanyang Technological University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8856-756X","authenticated-orcid":false,"given":"Mingkui","family":"Tan","sequence":"additional","affiliation":[{"name":"South China University of Technology","place":["Guangzhou, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00813"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00916"},{"key":"ref4","first-page":"1","article-title":"Benchmarking neural network robustness to common corruptions and perturbations","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hendrycks","year":"2019"},{"key":"ref5","first-page":"5637","article-title":"Wilds: A benchmark of in-the-wild distribution shifts","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Koh","year":"2021"},{"key":"ref6","first-page":"5389","article-title":"Do ImageNet classifiers generalize to ImageNet?","volume-title":"Proc. Int. Conf. Mach. Learn.","volume":"97","author":"Recht"},{"key":"ref7","first-page":"9229","article-title":"Test-time training with self-supervision for generalization under distribution shifts","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Sun","year":"2020"},{"key":"ref8","first-page":"1","article-title":"Tent: Fully test-time adaptation by entropy minimization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Wang","year":"2021"},{"key":"ref9","first-page":"2427","article-title":"Test-time classifier adjustment module for model-agnostic domain generalization","volume-title":"Proc. Adv. Neural Inf. Proc. Syst.","volume":"34","author":"Iwasawa","year":"2021"},{"key":"ref10","first-page":"3080","article-title":"Mt3: Meta test-time training for self-supervised test-time adaption","volume-title":"Proc. Int. Conf. Artif. Intell. Statist.","author":"Bartler","year":"2022"},{"key":"ref11","first-page":"21808","article-title":"TTT : When does self-supervised test-time training fail or thrive?","volume-title":"Proc. Adv. Neural Inf. Proc. Syst.","volume":"34","author":"Liu","year":"2021"},{"key":"ref12","first-page":"16888","article-title":"Efficient test-time model adaptation without forgetting","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Niu","year":"2022"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.5555\/3045118.3045167"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01261-8_1"},{"key":"ref15","article-title":"Layer normalization","author":"Ba","year":"2016"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-024-02213-5"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-024-02181-w"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2025.3560696"},{"key":"ref19","first-page":"46977","article-title":"Entropy is not enough for test-time adaptation: From the perspective of disentangled factors","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lee","year":"2024"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/WACV57701.2024.00254"},{"key":"ref21","article-title":"Generalizing across domains via cross-gradient training","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Shankar","year":"2018"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v32i1.11596"},{"key":"ref23","first-page":"6447","article-title":"Domain generalization via model-agnostic learning of semantic features","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Dou","year":"2019"},{"key":"ref24","article-title":"Robustness properties of facebook\u2019s ResNeXt WSL models","author":"Orhan","year":"2019"},{"key":"ref25","first-page":"6665","article-title":"Fast autoaugment","volume-title":"Proc. Adv. Neural Inf. Proc. Syst.","volume":"32","author":"Lim","year":"2019"},{"key":"ref26","article-title":"Augmix: A simple data processing method to improve robustness and uncertainty","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hendrycks","year":"2020"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01220"},{"key":"ref28","first-page":"25407","article-title":"Improving out-of-distribution robustness via selective augmentation","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Yao","year":"2022"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2024.3354420"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v32i1.11767"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00392"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2020.3006377"},{"key":"ref33","article-title":"COVID-DA: Deep domain adaptation from typical pneumonia to COVID-19","author":"Zhang","year":"2020"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19827-4_21"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00966"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00460"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2021\/402"},{"key":"ref38","first-page":"6028","article-title":"Do we really need to access the source data? source hypothesis transfer for unsupervised domain adaptation","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Liang","year":"2020"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3103390"},{"key":"ref40","first-page":"1","article-title":"Unsupervised representation learning by predicting image rotations","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Gidaris","year":"2018"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.5555\/3524938.3525087"},{"key":"ref42","article-title":"MECTA: Memory-economic continual test-time model adaptation","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hong","year":"2023"},{"key":"ref43","first-page":"11539","article-title":"Improving robustness against common corruptions by covariate shift adaptation","volume-title":"Proc. Adv. Neural Inf. Proc. Syst.","volume":"33","author":"Schneider","year":"2020"},{"key":"ref44","article-title":"Mixnorm:Test-time adaptation through online normalization estimation","author":"Hu","year":"2021"},{"key":"ref45","article-title":"Sita: Single image test-time adaptation","author":"Khurana","year":"2021"},{"key":"ref46","article-title":"TTN: A domain-shift aware batch normalization in test-time adaptation","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lim","year":"2023"},{"key":"ref47","article-title":"DELTA: Degradation-free fully test-time adaptation","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Zhao","year":"2023"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2799"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00706"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00039"},{"key":"ref51","first-page":"1","article-title":"Towards stable test-time adaptation in dynamic wild world","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Niu","year":"2023"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-024-02181-w"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0620"},{"key":"ref54","first-page":"24371","article-title":"Beyond entropy: Region confidence proxy for wild test-time adaptation","volume-title":"Proc. Int. Conf. Mach. Learn.","volume":"267","author":"Hu","year":"2025"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01528"},{"key":"ref56","article-title":"Generalized robust test-time adaptation in continuous dynamic scenarios","author":"Li","year":"2023"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1735"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.52202\/085713-2705"},{"key":"ref59","article-title":"Sharpness-aware minimization for efficiently improving generalization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Foret","year":"2021"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.1.1"},{"key":"ref61","first-page":"5905","article-title":"ASAM: Adaptive sharpness-aware minimization for scale-invariant learning of deep neural networks","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Kwon","year":"2021"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00806"},{"key":"ref63","article-title":"Efficient sharpness-aware minimization for improved training of neural networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Du","year":"2022"},{"key":"ref64","article-title":"When vision transformers outperform resnets without pre-training or strong data augmentations","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Chen","year":"2021"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2015509117"},{"key":"ref66","article-title":"Neural collapse under MSE loss: Proximity to and dynamics on the central path","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Han","year":"2021"},{"key":"ref67","article-title":"Neural collapse in test-time adaptation","author":"Chen","year":"2025"},{"key":"ref68","article-title":"Understanding dimensional collapse in contrastive self-supervised learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Jing","year":"2022"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00946"},{"key":"ref70","first-page":"12310","article-title":"Barlow twins: Self-supervised learning via redundancy reduction","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zbontar","year":"2021"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2130"},{"key":"ref72","article-title":"Evaluating prediction-time batch normalization for robustness under covariate shift","author":"Nado","year":"2020"},{"key":"ref73","article-title":"Rethinking batch in batchnorm","author":"Wu","year":"2021"},{"key":"ref74","article-title":"VICReg: Variance-invariance-covariance regularization for self-supervised learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Bardes","year":"2022"},{"key":"ref75","article-title":"Pytorch image models","author":"Wightman","year":"2019"},{"key":"ref76","article-title":"Back to the source: Diffusion-driven test-time adaptation","volume-title":"Proc. ICML Workshop Updatable Mach. Learn.","author":"Gao","year":"2022"},{"key":"ref77","first-page":"39578","article-title":"Let\u2019s verify step by step","volume-title":"Proc. 12th Int. Conf. Learn. Representations","author":"Lightman","year":"2024"},{"key":"ref78","first-page":"47885","article-title":"Mathscale: Scaling instruction tuning for mathematical reasoning","volume-title":"Proc. Int. Conf. Mach. Learn.","volume":"235","author":"Tang","year":"2024"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.52202\/068431-0278"},{"key":"ref80","first-page":"24823","article-title":"Test-time learning for large language models","volume-title":"Proc. Int. Conf. Mach. Learn.","volume":"267","author":"Hu","year":"2025"},{"key":"ref81","first-page":"1558","article-title":"Learning discrete representations via information maximizing self-augmented training","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Hu","year":"2017"},{"key":"ref82","first-page":"6391","article-title":"Visualizing the loss landscape of neural nets","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Li","year":"2018"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00823"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19827-4_26"},{"key":"ref86","first-page":"144893","article-title":"Zerosiam: An efficient asymmetry for test-time entropy optimization without collapse","volume-title":"Proc. Int. Conf. Learn. Representations","volume":"2026","author":"Chen"},{"key":"ref87","article-title":"Decoupled weight decay regularization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Loshchilov","year":"2019"},{"key":"ref88","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01167"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/34\/11723300\/11644824.pdf?arnumber=11644824","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T19:14:25Z","timestamp":1791486865000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11644824\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,11]]},"references-count":88,"journal-issue":{"issue":"11"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2026.3721895","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":[[2026,11]]}}}