FrameFusion: Combining Similarity and Importance for Video Token Reduction on Large Visual Language Models

Add code
Dec 30, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: