logo
CanyamAI summaries for academic research
language
MobileMobile
  • Phone FeatureWeChat Mini Program
EnglishEnglish
noticeNotice
BlogBlog
delete
arrow
Return

Memory augmented using diffusion model for class-incremental learning

delete2025-06-10
delete0
delete
OA
AI
Q
Quentin Jodelet
X
Xin Liu
Y
Yin Jun Phua
T
Tsuyoshi Murata
DOI：10.1016/j.imavis.2025.105600delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Generates synthetic samples of past classes using pre-trained diffusion models. • Synthetic data can be used for replay in supervised losses and for distillation loss. • Significantly improves performances of methods for class-incremental learning.
Keywords:
Continual learning
Class incremental learning
Image classification
Image generation

Journal

Image and Vision Computing cover
Image and Vision Computing
IF:
4.2
Papers:
4.0K
Citations:
6.7K

Organization

No organization information available