Why Picking 150 Frames Determines Whether an LLM Actually Watches Video
Vision‑LLMs can only ingest about 150 images per clip, so the frames you keep decide if the model sees the action or just a slideshow. Uniform sampling wastes tokens on redundant shots and skips critical moments; a scene‑score plus rolling‑average filter picks frames that earn their slot. This boosts video understanding without blowing the token budget.