AI Paper Atlas

Research map / Video / Video-language understanding

Contribution-aware Token Compression for Efficient Video Understanding via Reinforcement Learning

Yinchao Ma, Qiang Zhou, Zhibin Wang, Xianing Chen, Hanqing Yang, Jun Song, Bo Zheng

AAAI 2026 · 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
use OCSS instead of random sampling in policy training (ablation)CaCoVID (LLaVA-OV-7B) · VideoMMEaccuracy: 54.4 -> 58.5Improved
use OCSS instead of multinomial sampling (ablation)CaCoVID · VideoMMEaccuracy: 55.1 -> 58.5Improved
sample within each frame rather than over whole video (ablation)CaCoVID · VideoMMEaccuracy: 57.2 -> 58.5Improved
raise training sampling ratio from 2% to 5% (ablation)CaCoVID · VideoMMEaccuracy: 58.5 -> 57.3Hurt
lower training sampling ratio from 2% to 1% (ablation)CaCoVID · VideoMMEaccuracy: 58.5 -> 58.3No clear effect
add ineffective sample filtering (ablation)CaCoVID · VideoMMEaccuracy: 57.1 -> 57.3Improved
add experience replay (ablation)CaCoVID · VideoMMEaccuracy: 57.3 -> 58.1Improved
add dynamic sample ratio (ablation)CaCoVID · VideoMMEaccuracy: 58.1 -> 58.5Improved
adaptive per-frame token budget instead of equal (ablation)CaCoVID · VideoMMEaccuracy: 57.4 -> 58.1Improved
add uniform spatio-temporal tokens to retained set (ablation)CaCoVID · VideoMMEaccuracy: 58.1 -> 58.5Improved
use CaCoVID instead of DivPrune/PruneVID for compression (ablation)LLaVA-OneVision-7B 25% · avg accuracyaccuracy: 55.0 -> 55.8Improved
use CaCoVID instead of PruneVID at 10% retentionQwen2.5-VL-3B · avg accuracyaccuracy: 50.2 -> 52.0Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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