MoE-Adapters++ : Towards More Efficient Continual Learning of Vision-Language Models via Dynamic Mixture-of-Experts Adapters
In this paper, we first propose MoE-Adapters, a parameter-efficient training framework to alleviate long-term forgetting issues in incremental learning with Vision-Language Models (VLM). Our MoE-Adapters leverages incrementally added routers to activate and integrate exclusive expert adapters from a...
Ausführliche Beschreibung
Bibliographische Detailangaben
Veröffentlicht in: | IEEE transactions on pattern analysis and machine intelligence. - 1979. - PP(2025) vom: 11. Aug.
|
1. Verfasser: |
Yu, Jiazuo
(VerfasserIn) |
Weitere Verfasser: |
Huang, Zichen,
Zhuge, Yunzhi,
Zhang, Lu,
Hu, Ping,
Wang, Dong,
Lu, Huchuan,
He, You |
Format: | Online-Aufsatz
|
Sprache: | English |
Veröffentlicht: |
2025
|
Zugriff auf das übergeordnete Werk: | IEEE transactions on pattern analysis and machine intelligence
|
Schlagworte: | Journal Article |