EKPC reduces forgetting in class-incremental learning by selectively regularizing shared adapter parameters according to their importance and by compensating old class prototypes using a trainable semantic drift estimate.
In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 8218--8227
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
EKPC: Elastic Knowledge Preservation and Compensation for Class-Incremental Learning
EKPC reduces forgetting in class-incremental learning by selectively regularizing shared adapter parameters according to their importance and by compensating old class prototypes using a trainable semantic drift estimate.