STAR: Stage-Wise Attention-Guided Token Reduction for Efficient Large Vision-Language Models Inference

Add code
May 18, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: