AI Paper Atlas

Research map / Video / Video-language understanding

AV-Reasoner: Improving and Benchmarking Clue-Grounded Audio-Visual Counting for MLLMs

Lidong Lu, Guo Chen, Zhu Wei, Zhiqi Li, Yicheng Liu, Tong Lu

CVPR 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 14 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
train with GRPO curriculum (AV-Reasoner) instead of baseOla-Omni-7B · DVD-Countingaccuracy: 16.50 -> 43.50Improved
train with GRPO curriculumOla-Omni-7B · CG-AV-Counting (Long Acc)accuracy: 17.92 -> 22.30Improved
add SFT only (ablation)Ola-Omni-7B · DVD-Countingaccuracy: 16.50 -> 41.50Improved
add SFT only (ablation)Ola-Omni-7B · CG-AV-Counting (Long Acc)accuracy: 17.92 -> 15.00Hurt
add QA-based RL after SFT (ablation)SFT model · DVD-Countingaccuracy: 41.50 -> 23.00Hurt
add grounding RL (ablation)SFT + QA RL · CG-AV-Counting (Long Acc)accuracy: 16.55 -> 18.21Improved
add counting RL (ablation)SFT + QA + grounding RL · CG-AV-Counting (Long Acc)accuracy: 18.21 -> 20.84Improved
add full-task RL (ablation)CB-RL with SRM · CG-AV-Counting (Long Acc)accuracy: 20.84 -> 21.03Improved
curriculum RL instead of GRPO on all data at once (ablation)SFT model · CG-AV-Counting (Long Acc)accuracy: 10.42 -> 20.84Improved
add Stage Review Mechanism (ablation)CB-RL + FT-RL · CG-AV-Counting (Long Acc)accuracy: 20.74 -> 21.03No clear effect
add Stage Review Mechanism (ablation)CB-RL + FT-RL · LLP segment-levelaccuracy: 61.30 -> 64.70Mixed
enable explicit thinking output at inference (ablation)AV-Reasoner · AVHBench A2Vaccuracy: 84.45 -> 82.45Hurt

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