Continual learning in the presence of repetition
Journal
Neural Networks
ISSN
0893-6080
Type
journal article
Date Issued
2025-03
Author(s)
;
Pellegrini, Lorenzo
;
Duan, Xiaotian
;
Zhao, Zixuan
;
Xia, Fangfang
;
Masana, Marc
;
Tscheschner, Benedikt
;
Veas, Eduardo
;
Zheng, Yuxiang
;
Zhao, Shiji
;
Li, Shao-Yuan
;
Huang, Sheng-Jun
;
Lomonaco, Vincenzo
;
van de Ven, Gido M.
Abstract
Continual learning (CL) provides a framework for training models in ever-evolving environments. Although re-occurrence of previously seen objects or tasks is common in real-world problems, the concept of repetition in the data stream is not often considered in standard benchmarks for CL. Unlike with the rehearsal mechanism in buffer-based strategies, where sample repetition is controlled by the strategy, repetition in the data stream naturally stems from the environment. This report provides a summary of the CLVision challenge at CVPR 2023, which focused on the topic of repetition in class-incremental learning. The report initially outlines the challenge objective and then describes three solutions proposed by finalist teams that aim to effectively exploit the repetition in the stream to learn continually. The experimental results from the challenge highlight the effectiveness of ensemble-based solutions that employ multiple versions of similar modules, each trained on different but overlapping subsets of classes. This report underscores the transformative potential of taking a different perspective in CL by employing repetition in the data stream to foster innovative strategy design.
Language
English
Keywords
Continual learning
Class-incremental learning
Repetition
Competition
HSG Classification
contribution to scientific community
Refereed
Yes
Publisher
Elsevier BV
Volume
183
Start page
Art. 106920
Subject(s)
Division(s)
File(s)![Thumbnail Image]()
Name
1-s2.0-S0893608024008499-main.pdf
Type
Main Article
Size
2.04 MB
Format
Adobe PDF
Checksum (MD5)
c1cf89501b58805d925feedf23f6c138