GitHub / thu-nics / FrameFusion

The official code implementation of paper "Combining Similarity and Importance for Video Token Reduction on Large Visual Language Models"

JSON API: http://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/thu-nics%2FFrameFusion
PURL: pkg:github/thu-nics/FrameFusion

Stars: 46
Forks: 1
Open issues: 0

License: mit
Language: Python
Size: 19.9 MB
Dependencies parsed at: Pending

Created at: 7 months ago
Updated at: 13 days ago
Pushed at: 13 days ago
Last synced at: 12 days ago

Topics: efficient-deep-learning, llm, lvlm, video

Loading...