AI Paper Atlas

Research map / Video / Video-language understanding

Video2Reaction: Mapping Video to Audience Reaction Distribution in the Wild

Trang Nguyen, Sidong Zhang, Shiv Shankar, Gauri Jagatap, Deepak Chandran, Andrea Fanelli, Madalina Fiterau

ECCV 2026 · Video › Video-language understanding · approach: Event and anomaly reasoning in video · 1 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
LoRA finetuning on Video2ReactionLLaVA-Next-Video-7B · Video2ReactionChebyshev: 0.4110 -> 0.1882Improved
LoRA finetuning on Video2ReactionLLaVA-Next-Video-7B · Video2ReactionTop-3 F1: 0.1374 -> 0.7672Improved
LoRA finetuning on Video2ReactionLLaVA-Next-Video-7B · Video2ReactionKL divergence: 1.5547 -> 3.1765Hurt
LoRA finetuning on Video2ReactionQwen2.5-VL · Video2Reactioncosine: 0.4401 -> 0.8427Improved
LoRA finetuning on Video2ReactionQwen2.5-VL · Video2ReactionKL divergence: 1.5216 -> 3.4431Hurt
add text (clip description) modality (ablation)SA-BFGS · Video2ReactionMRR: 0.552 -> 0.709Improved
add audio on top of visual + text (ablation)SA-BFGS · Video2ReactionMRR: 0.709 -> 0.716No clear effect
add text to visual input (ablation)LLaVA-Next · Video2ReactionMRR: 0.731 -> 0.783Improved
two-stage (rephrase then label) annotation instead of single-stage (ablation)LLM annotation pipeline · human-LLM correlationmean Spearman correlation: 0.34 -> 0.40Improved
fine-tune on Video2Reaction plus 1% VCE vs 10% in-domain VCE onlyQwen2.5-VL · VCE benchmarkTop-3 accuracy: 0.35 -> 0.46Improved
specialized LDL (SA-BFGS) vs finetuned VLM on dominant reactionsSA-BFGS · Video2ReactionTop-1 F1: 0.6521 (LLaVA FT) vs 0.5283Hurt

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