Memory-efficient Continual Learning with Prototypical Exemplar Condensation

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Nguyen, Minh-Duong, Dao, Thien-Thanh, Nguyen, Le-Tuan, Le, Dung D., Wong, Kok-Seng
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913020607725568
author Nguyen, Minh-Duong
Dao, Thien-Thanh
Nguyen, Le-Tuan
Le, Dung D.
Wong, Kok-Seng
author_facet Nguyen, Minh-Duong
Dao, Thien-Thanh
Nguyen, Le-Tuan
Le, Dung D.
Wong, Kok-Seng
contents Rehearsal-based continual learning (CL) mitigates catastrophic forgetting by maintaining a subset of samples from previous tasks for replay. Existing studies primarily focus on optimizing memory storage through coreset selection strategies. While these methods are effective, they typically require storing a substantial number of samples per class (SPC), often exceeding 20, to maintain satisfactory performance. In this work, we propose to further compress the memory footprint by synthesizing and storing prototypical exemplars, which can form representative prototypes when passed through the feature extractor. Owing to their representative nature, these exemplars enable the model to retain previous knowledge using only a small number of samples while preserving privacy. Moreover, we introduce a perturbation-based augmentation mechanism that generates synthetic variants of previous data during training, thereby enhancing CL performance. Extensive evaluations on widely used benchmark datasets and settings demonstrate that the proposed algorithm achieves superior performance compared to existing baselines, particularly in scenarios involving large-scale datasets and a high number of tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2603_13804
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Memory-efficient Continual Learning with Prototypical Exemplar Condensation
Nguyen, Minh-Duong
Dao, Thien-Thanh
Nguyen, Le-Tuan
Le, Dung D.
Wong, Kok-Seng
Machine Learning
Artificial Intelligence
68T05
I.2.6
Rehearsal-based continual learning (CL) mitigates catastrophic forgetting by maintaining a subset of samples from previous tasks for replay. Existing studies primarily focus on optimizing memory storage through coreset selection strategies. While these methods are effective, they typically require storing a substantial number of samples per class (SPC), often exceeding 20, to maintain satisfactory performance. In this work, we propose to further compress the memory footprint by synthesizing and storing prototypical exemplars, which can form representative prototypes when passed through the feature extractor. Owing to their representative nature, these exemplars enable the model to retain previous knowledge using only a small number of samples while preserving privacy. Moreover, we introduce a perturbation-based augmentation mechanism that generates synthetic variants of previous data during training, thereby enhancing CL performance. Extensive evaluations on widely used benchmark datasets and settings demonstrate that the proposed algorithm achieves superior performance compared to existing baselines, particularly in scenarios involving large-scale datasets and a high number of tasks.
title Memory-efficient Continual Learning with Prototypical Exemplar Condensation
topic Machine Learning
Artificial Intelligence
68T05
I.2.6
url https://arxiv.org/abs/2603.13804