{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,13]],"date-time":"2026-01-13T05:12:29Z","timestamp":1768281149852,"version":"3.49.0"},"reference-count":84,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"2","license":[{"start":{"date-parts":[[2026,2,1]],"date-time":"2026-02-01T00:00:00Z","timestamp":1769904000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,2,1]],"date-time":"2026-02-01T00:00:00Z","timestamp":1769904000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,2,1]],"date-time":"2026-02-01T00:00:00Z","timestamp":1769904000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"NTU RSR"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2026,2]]},"DOI":"10.1109\/tpami.2025.3619883","type":"journal-article","created":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T17:36:45Z","timestamp":1760117805000},"page":"1811-1824","source":"Crossref","is-referenced-by-count":0,"title":["Release the Potential of Memory Buffer in Continual Learning: A Dynamic System Perspective"],"prefix":"10.1109","volume":"48","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-2780-9446","authenticated-orcid":false,"given":"Zhenyi","family":"Wang","sequence":"first","affiliation":[{"name":"Department of Computer Science and Institute of Artificial Intelligence, University of Central Florida, Orlando, FL, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5659-3464","authenticated-orcid":false,"given":"Li","family":"Shen","sequence":"additional","affiliation":[{"name":"School of Cyber Science and Technology, Shenzhen Campus of Sun Yat-sen University, Shenzhen, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4323-642X","authenticated-orcid":false,"given":"Tiehang","family":"Duan","sequence":"additional","affiliation":[{"name":"Department of Computer Science, College of Computing, Grand Valley State University, Allendale, MI, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8841-6072","authenticated-orcid":false,"given":"Yanjun","family":"Zhu","sequence":"additional","affiliation":[{"name":"Northeastern University, Boston, MA, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9640-6472","authenticated-orcid":false,"given":"Tongliang","family":"Liu","sequence":"additional","affiliation":[{"name":"The University of Sydney, Camperdown, NSW, Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5488-8514","authenticated-orcid":false,"given":"Mingchen","family":"Gao","sequence":"additional","affiliation":[{"name":"University at Buffalo, Getzville, NY, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7225-5449","authenticated-orcid":false,"given":"Dacheng","family":"Tao","sequence":"additional","affiliation":[{"name":"College of Computing and Data Science, and Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","first-page":"11849","article-title":"Online continual learning with maximal interfered retrieval","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"32","author":"Aljundi","year":"2019"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.01151"},{"key":"ref3","article-title":"Gradient based sample selection for online continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Aljundi","year":"2019"},{"key":"ref4","article-title":"Learning fast, learning slow: A general continual learning method based on complementary learning system","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Arani","year":"2022"},{"issue":"Nov","key":"ref5","first-page":"463","article-title":"Rademacher and Gaussian complexities: Risk bounds and structural results","volume":"3","author":"Bartlett","year":"2002","journal-title":"J. Mach. Learn. Res."},{"key":"ref6","first-page":"31886","article-title":"On the effectiveness of lipschitz-driven rehearsal in continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"35","author":"Bonicelli","year":"2022"},{"key":"ref7","article-title":"Dark experience for general continual learning: A strong, simple baseline","volume-title":"Proc. 34th Conf. Neural Inf. Process. Syst.","author":"Buzzega","year":"2020"},{"key":"ref8","article-title":"New insights on reducing abrupt representation change in online continual learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Caccia","year":"2022"},{"key":"ref9","first-page":"1240","article-title":"Online learned continual compression with adaptive quantization modules","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Caccia","year":"2020"},{"key":"ref10","article-title":"CPR: Classifier-projection regularization for continual learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Cha","year":"2021"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v35i8.16861"},{"key":"ref12","article-title":"Efficient lifelong learning with A-GEM","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Chaudhry","year":"2019"},{"key":"ref13","article-title":"Continual learning with tiny episodic memories","author":"Chaudhry","year":"2019"},{"key":"ref14","article-title":"Mode regularized generative adversarial networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Che","year":"2017"},{"key":"ref15","article-title":"Neural ordinary differential equations","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"31","author":"Chen","year":"2018"},{"issue":"1","key":"ref16","first-page":"9885","article-title":"A group-theoretic framework for data augmentation","volume":"21","author":"Chen","year":"2020","journal-title":"J. Mach. Learn. Res."},{"key":"ref17","first-page":"1952","article-title":"Online continual learning from imbalanced data","volume-title":"Proc. 37th Int. Conf. Mach. Learn.","volume":"119","author":"Chrysakis","year":"2020"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3057446"},{"key":"ref19","article-title":"Uncertainty-guided continual learning with Bayesian neural networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Ebrahimi","year":"2020"},{"key":"ref20","article-title":"PathNet: Evolution channels gradient descent in super neural networks","author":"Fernando","year":"2017"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1007\/BF00344251"},{"key":"ref22","first-page":"10744","article-title":"DDGR: Continual learning with deep diffusion-based generative replay","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Gao","year":"2023"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01144"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW50498.2020.00118"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref26","article-title":"Task agnostic continual learning via meta learning","author":"He","year":"2019"},{"key":"ref27","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":"ref28","first-page":"14135","article-title":"Posterior meta-replay for continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Henning","year":"2021"},{"key":"ref29","first-page":"29193","article-title":"Gradient-based editing of memory examples for online task-free continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Jin","year":"2021"},{"key":"ref30","first-page":"18747","article-title":"Efficient and accurate gradients for neural SDEs","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Kidger","year":"2021"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1611835114"},{"key":"ref32","article-title":"Learning multiple layers of features from tiny images","author":"Krizhevsky","year":"2009"},{"key":"ref33","article-title":"Improved precision and recall metric for assessing generative models","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Kynk\u00e4\u00e4nniemi","year":"2019"},{"key":"ref34","article-title":"A neural dirichlet process mixture model for task-free continual learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lee","year":"2020"},{"key":"ref35","first-page":"3870","article-title":"Scalable gradients for stochastic differential equations","volume-title":"Proc. Int. Conf. Artif. Intell. Statist.","author":"Li","year":"2020"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2017.2773081"},{"key":"ref37","article-title":"Continual learning with recursive gradient optimization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Liu","year":"2022"},{"key":"ref38","article-title":"Neural SDE: Stabilizing neural ode networks with stochastic noise","author":"Liu","year":"2019"},{"key":"ref39","article-title":"Gradient episodic memory for continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Lopez-Paz","year":"2017"},{"key":"ref40","first-page":"3276","article-title":"Beyond finite layer neural networks: Bridging deep architectures and numerical differential equations","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Lu","year":"2018"},{"key":"ref41","article-title":"Continual learning with filter atom swapping","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Miao","year":"2022"},{"key":"ref42","article-title":"Variational continual learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Nguyen","year":"2018"},{"key":"ref43","volume-title":"Stochastic Differential Equations: An Introduction With Applications (Universitext)","author":"\u00d8ksendal","year":"2014"},{"key":"ref44","first-page":"5250","article-title":"Learning where to learn: Gradient sparsity in meta and continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Von","year":"2021"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1201\/9780203749319"},{"key":"ref46","article-title":"Looking back on learned experiences for class\/task incremental learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"PourKeshavarzi","year":"2022"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00360"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58536-5_31"},{"key":"ref49","first-page":"20608","article-title":"BNS: Building network structures dynamically for continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Qin","year":"2021"},{"key":"ref50","first-page":"17284","article-title":"Formalizing the generalization-forgetting trade-off in continual learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Raghavan","year":"2021"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.587"},{"key":"ref52","article-title":"Learning to learn without forgetting by maximizing transfer and minimizing interference","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Riemer","year":"2019"},{"key":"ref53","article-title":"Progressive neural networks","author":"Rusu","year":"2016"},{"key":"ref54","article-title":"Gradient projection memory for continual learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Saha","year":"2021"},{"key":"ref55","article-title":"Probabilistic ODE solvers with Runge-Kutta means","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"27","author":"Schober","year":"2014"},{"key":"ref56","first-page":"4528","article-title":"Progress and compress: A scalable framework for continual learning","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Schwarz","year":"2018"},{"key":"ref57","first-page":"4548","article-title":"Overcoming catastrophic forgetting with hard attention to the task","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Serr\u00e1","year":"2018"},{"key":"ref58","article-title":"Continual learning with deep generative replay","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Shin","year":"2017"},{"key":"ref59","article-title":"VEEGAN: Reducing mode collapse in GANs using implicit variational learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Srivastava","year":"2017"},{"key":"ref60","article-title":"Tiny imagenet challenge (cs231n)","year":"2015"},{"key":"ref61","first-page":"67794","article-title":"COOM: A game benchmark for continual reinforcement learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"36","author":"Tomilin","year":"2023"},{"key":"ref62","article-title":"Three scenarios for continual learning","author":"van de Ven","year":"2019"},{"key":"ref63","article-title":"Matching networks for one shot learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"29","author":"Vinyals","year":"2016"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1145\/3147.3165"},{"key":"ref65","article-title":"Continual learning with hypernetworks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"von Oswald","year":"2020"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3367329"},{"key":"ref67","article-title":"Memory replay with data compression for continual learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Wang","year":"2022"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.52202\/079017-4215"},{"key":"ref69","article-title":"A unified and general framework for continual learning","volume-title":"Proc. 12th Int. Conf. Learn. Representations","author":"Wang","year":"2024"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3317874"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00782"},{"key":"ref72","first-page":"22985","article-title":"Improving task-free continual learning by distributionally robust memory evolution","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Wang","year":"2022"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02349"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3498346"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00024"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00046"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v36i3.20198"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-20083-0_6"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01928"},{"key":"ref80","article-title":"Lifelong learning with dynamically expandable networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Yoon","year":"2018"},{"key":"ref81","first-page":"3987","article-title":"Continual learning through synaptic intelligence","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zenke","year":"2017"},{"key":"ref82","first-page":"3987","article-title":"Continual learning through synaptic intelligence","volume-title":"Proc. 34th Int. Conf. Mach. Learn.","volume":"70","author":"Zenke","year":"2017"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.1162\/neco_a_01430"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00581"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/34\/11345188\/11198838.pdf?arnumber=11198838","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,12]],"date-time":"2026-01-12T22:00:37Z","timestamp":1768255237000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11198838\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2]]},"references-count":84,"journal-issue":{"issue":"2"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2025.3619883","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":[[2026,2]]}}}