AVAM: Universal Training-free Adaptive Visual Anchoring Embedded into Multimodal Large Language Model for Multi-image Question Answering

Add code
Aug 25, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: