LLaVA-Scissor introduces a training-free token compression method for video multimodal models, utilizing Semantic Connected Components for effective, non-redundant semantic coverage, outperforming existing methods in various benchmarks.

https://arxiv.org/abs//2506.21862

YouTube: https://www.youtube.com/@ArxivPapers

TikTok: https://www.tiktok.com/@arxiv_papers

Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id1692476016

Spotify: https://podcasters.spotify.com/pod/show/arxiv-papers