
An efficient pipeline, RED-VILLM, is proposed to quickly develop high-performing Video Large Language Models (Video LLMs) by leveraging the foundational work of Image Large Language Models (Image LLMs) and incorporating a plug-and-play temporal adaptation module.


coremsg

efficient-transition-from-image-to-video-large-language-models-leveraging-priors-and-temporal-adaptation


Efficient Transition from Image to Video Large Language Models: Leveraging Priors and Temporal Adaptation


title_rewrite


SlowFast-LLaVA is a training-free video large language model that can effectively capture both detailed spatial semantics and long-range temporal context without exceeding the token budget of commonly used language models.


a-strong-training-free-baseline-for-video-large-language-models-slowfast-llava


A Strong Training-Free Baseline for Video Large Language Models: SlowFast-LLaVA
