[ICCV2023] - CTP: Towards Vision-Language Continual Pretraining via Compatible Momentum Contrast and Topology Preservation
Geek Repo:Geek Repo
Github PK Tool:Github PK Tool