Picture for Xiangfeng Xu

Xiangfeng Xu

Dynamic-LLaVA: Efficient Multimodal Large Language Models via Dynamic Vision-language Context Sparsification

Add code
Dec 03, 2024
Viaarxiv icon