Prompt-based learning for few-shot class-incremental learning

Few-Shot Class-Incremental Learning (FSCIL) aims to enable deep neural networks to incrementally learn new tasks from a limited number of labeled samples, while retaining knowledge of previously learned tasks, mimicking the way humans learn. In this paper, we introduce a novel approach called Prompt...

Celý popis

Podrobná bibliografie

Hlavní autoři:	Jicheng Yuan, Hang Chen, Songsong Tian, Wenfa Li, Lusi Li, Enhao Ning, Yugui Zhang
Médium:	Artigo
Jazyk:	Inglês
Vydáno:	Elsevier 2025-05-01
Edice:	Alexandria Engineering Journal
Témata:	Few-shot class-incremental learning Prompt learning Prototype classifier Catastrophic forgetting
On-line přístup:	http://www.sciencedirect.com/science/article/pii/S1110016825001735
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!