{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,7]],"date-time":"2026-08-07T19:47:36Z","timestamp":1786132056534,"version":"3.56.0"},"reference-count":78,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"9","license":[{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"Beijing Natural Science Foundation","award":["L233036"],"award-info":[{"award-number":["L233036"]}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["62373343"],"award-info":[{"award-number":["62373343"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2026,9]]},"DOI":"10.1109\/tpami.2026.3683792","type":"journal-article","created":{"date-parts":[[2026,4,14]],"date-time":"2026-04-14T19:35:50Z","timestamp":1776195350000},"page":"10221-10234","source":"Crossref","is-referenced-by-count":8,"title":["Neural Network Optimization Reimagined: Decoupled Techniques for Scratch and Fine-Tuning"],"prefix":"10.1109","volume":"48","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7897-1673","authenticated-orcid":false,"given":"Xin","family":"Ning","sequence":"first","affiliation":[{"name":"Chinese Academy of Sciences","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5121-1682","authenticated-orcid":false,"given":"Qiankun","family":"Li","sequence":"additional","affiliation":[{"name":"Nanyang Technological University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-5502-2535","authenticated-orcid":false,"given":"Xiaolong","family":"Huang","sequence":"additional","affiliation":[{"name":"Mila-Quebec AI Institute","place":["Montreal, QC, Canada"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1036-6762","authenticated-orcid":false,"given":"Qiupu","family":"Chen","sequence":"additional","affiliation":[{"name":"University of Science and Technology of China","place":["Hefei, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5059-4423","authenticated-orcid":false,"given":"Feng","family":"He","sequence":"additional","affiliation":[{"name":"University of Science and Technology of China","place":["Hefei, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9668-2883","authenticated-orcid":false,"given":"Weijun","family":"Li","sequence":"additional","affiliation":[{"name":"Chinese Academy of Sciences","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2851-4260","authenticated-orcid":false,"given":"Prayag","family":"Tiwari","sequence":"additional","affiliation":[{"name":"Halmstad University","place":["Halmstad, Sweden"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9066-1475","authenticated-orcid":false,"given":"Xinwang","family":"Liu","sequence":"additional","affiliation":[{"name":"National University of Defense Technology","place":["Changsha, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/JBHI.2023.3343518"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v38i11.29159"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1833"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-015-0816-y"},{"key":"ref5","article-title":"Laion-400 M: Open dataset of clip-filtered 400 million image-text pairs","author":"Schuhmann","year":"2021"},{"key":"ref6","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Radford"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01553"},{"key":"ref8","article-title":"Beit: BERT pre-training of image transformers","author":"Bao","year":"2021"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3129870"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3357503"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3581783.3612496"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/3664647.3688975"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2025.111511"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3237935"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3435571"},{"key":"ref16","first-page":"11448","article-title":"Positive-negative momentum: Manipulating stochastic gradient noise to improve generalization","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Xie"},{"key":"ref17","article-title":"Rotational equilibrium: How weight decay balances learning across neural networks","author":"Kosson","year":"2024"},{"key":"ref18","first-page":"177","article-title":"Comparing biases for minimal network construction with back-propagation","volume-title":"Proc. 2nd Int. Conf. Neural Inf. Process. Syst.","author":"Hanson"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.52202\/075280-3458"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3057446"},{"key":"ref21","article-title":"Improving generalization performance by switching from adam to SGD","author":"Keskar","year":"2017"},{"key":"ref22","article-title":"Adam: A method for stochastic optimization","author":"Kingma","year":"2014"},{"key":"ref23","article-title":"Globally optimal training of neural networks with threshold activation functions","author":"Ergen","year":"2023"},{"key":"ref24","article-title":"Equi-normalization of neural networks","author":"Stock","year":"2019"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1038\/323533a0"},{"issue":"7","key":"ref26","first-page":"2121","article-title":"Adaptive subgradient methods for online learning and stochastic optimization","volume":"12","author":"Duchi","year":"2011","journal-title":"J. Mach. Learn. Res."},{"key":"ref27","first-page":"950","article-title":"A simple weight decay can improve generalization","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Krogh"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58452-8_13"},{"key":"ref29","article-title":"Decoupled weight decay regularization","author":"Loshchilov","year":"2017"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-33374-3_26"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.587"},{"key":"ref32","first-page":"350","article-title":"Experience replay for continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Rolnick"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW50498.2020.00121"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-13324-4_47"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1611835114"},{"key":"ref36","first-page":"2825","article-title":"Explicit inductive bias for transfer learning with convolutional networks","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Xuhong"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19827-4_41"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01900"},{"key":"ref39","article-title":"Learning multiple layers of features from tiny images","author":"Krizhevsky","year":"2009"},{"issue":"2","key":"ref40","first-page":"4","article-title":"Reading digits in natural images with unsupervised feature learning","volume-title":"Proc. NIPS Workshop Deep Learn. Unsupervised Feature Learn.","volume":"2011","author":"Netzer","year":"2011"},{"key":"ref41","article-title":"The Caltech-UCSD birds-200\u20132011 dataset","author":"Wah","year":"2011"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/ICCVW.2013.77"},{"key":"ref43","first-page":"487","article-title":"Learning deep features for scene recognition using places database","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Zhou"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-013-0695-z"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.572"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.591"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-10602-1_48"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-018-1140-0"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1038\/nature14539"},{"key":"ref50","article-title":"Language models are few-shot learners","author":"Brown","year":"2020"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-020-01316-z"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.emnlp-main.98"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.11"},{"key":"ref55","article-title":"Bert: Pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018"},{"issue":"140","key":"ref56","first-page":"1","article-title":"Exploring the limits of transfer learning with a unified text-to-text transformer","volume":"21","author":"Raffel","year":"2020","journal-title":"J. Mach. Learn. Res."},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i10.33204"},{"key":"ref58","article-title":"On the opportunities and risks of foundation models","author":"Bommasani","year":"2021"},{"key":"ref59","article-title":"Incremental classifier learning with generative adversarial networks","author":"Wu","year":"2018"},{"key":"ref60","article-title":"Re-evaluating continual learning scenarios: A categorization and case for strong baselines","author":"Hsu","year":"2018"},{"key":"ref61","first-page":"3987","article-title":"Continual learning through synaptic intelligence","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zenke"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2017.2773081"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.148"},{"key":"ref64","first-page":"2790","article-title":"Parameter-efficient transfer learning for NLP","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Houlsby"},{"key":"ref65","article-title":"LoRA: Low-rank adaptation of large language models","author":"Hu","year":"2021"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.eacl-main.39"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1214\/aoms\/1177729586"},{"key":"ref68","first-page":"1139","article-title":"On the importance of initialization and momentum in deep learning","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Sutskever"},{"key":"ref69","article-title":"On the variance of the adaptive learning rate and beyond","author":"Liu","year":"2019"},{"key":"ref70","article-title":"Adaptive gradient methods with dynamic bound of learning rate","author":"Luo","year":"2019"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0705"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.106"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.4324\/9781410605337-29"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01167"},{"key":"ref76","article-title":"An image is worth 16 \u00d7 16 words: Transformers for image recognition at scale","author":"Dosovitskiy","year":"2020"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.76"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/34\/11643729\/11481088.pdf?arnumber=11481088","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,7]],"date-time":"2026-08-07T19:09:24Z","timestamp":1786129764000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11481088\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9]]},"references-count":78,"journal-issue":{"issue":"9"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2026.3683792","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":[[2026,9]]}}}