AI Paper Atlas

Research map / Video / Video-language understanding

Learning Beyond Still Frames: Scaling Vision-Language Models with Video

Yiyuan Zhang, Handong Li, Jing Liu, Xiangyu Yue

ICCV 2025 · Video › Video-language understanding · approach: Long video understanding · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
scale video data in Stage I pretraining (ablation)LLaVA-Prime · VideoMME / Egoschemaaccuracy: 55.1 -> 57.3 (VideoMME)Improved
scale pretraining video data from 2M to 128M (ablation)LLaVA-Prime-7B · VideoMME generalscore (%): 56.2 -> 63.8Improved
scale pretraining video data from 2M to 128M (ablation)LLaVA-Prime-7B · general VL tasks (MME, MMBench, POPE, SEED, MMMU)score (%): 66.2 -> 72.2Improved
scale video data for long videos (ablation)LLaVA-Prime-7B · VideoMME longscore (%): 46.9 at 16M -> 55.5 at 128MMixed
increase video frames rate to 5 fps (ablation)LLaVA-Prime · VideoMMEaccuracy: 57.5 -> 59.8 (1 fps to 5 fps)Improved
increase video resolution from 224 to 336 (ablation)LLaVA-Prime · VideoMMEaccuracy: 55.1 -> 54.8Hurt
add video pretraining of LLM (Stage II)Qwen-2-7B · CMMLU / MMLU-Pro / MMLUaccuracy (%): CMMLU +3.4, MMLU-Pro +1.0, MMLU 68.2 -> 48.6Mixed
add Stage III causal hierarchical aggregation (ablation)LLaVA-Prime · VideoMMEaccuracy: 58.8 -> 60.2Improved
scale LLM size (ablation)LLaVA-Prime · VideoMMEaccuracy: 55.1 -> 62.8Improved
video-pretrain ViT-g vision encoder vs EVA-01-CLIP-g/14LLaVA-Prime-ViT-g · zero-shot classification (27 datasets)avg top-1 accuracy (%): 72.4 -> 78.5Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy