AI Paper Atlas

Research map / Video / Video generation

HumanScore: Benchmarking Human Motions in Generated Videos

Tiange Xiang, Yusu Fang, Tian Tan, Narayan Schuetz, Scott Delp, Fei-Fei Li, Ehsan Adeli

NeurIPS 2026 · Video › Video generation · approach: Trajectory-guided motion control · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Introduces HumanScore, a biomechanics-informed benchmark that scores human motion quality in AI-generated videos via six metrics (anatomy, kinematics, kinetics) and evaluates 13 video generators.

Problem

Existing video generation benchmarks emphasize visual realism and semantic alignment and do not systematically measure the biomechanical plausibility of human figures and motions in generated videos.

Approach

Extract 3D human pose/mesh and OpenSim skeletons from generated videos and compute six interpretable biomechanics-grounded metrics (extra limbs, bone-length stability, joint range of motion, self-collision, kinematic extremes, motion smoothness) aggregated via frequency, severity, and persistence, rather than relying on appearance or distributional metrics.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace MeTRAbs-based fitting with temporally optimized keypoints or PromptHMR for pose estimation (ablation)HumanScore metrics · model ranking robustnessmodel rankingNo clear effect
vary tolerance scales from strict to permissive (ablation)HumanScore metrics · model ranking robustnessmodel rankingNo clear effect
grid search over aggregation weights (alpha, beta, gamma) (ablation)HumanScore aggregation · Wan 2.2, CogVideoX-5B, Sora-2, Veo 3.1 fast and real videosmodel rankingNo clear effect
add systematic prompt refinement (studio, full-body, static camera, single person)video generators · HumanScore motion realismmotion realismNo clear effect
compare HumanScore model ranking with human preference win ratiosHumanScore · human preference study (~1200 responses)Spearman correlation: close to 1.0 across all evaluation dimensionsImproved
compare kinetic metrics against appearance-based VBench axesHumanScore Kinetic vs Kinematic · VBench correlationSpearman correlation: Kinematic 0.927 vs Kinetic 0.349 (imaging quality)Mixed
increase motion difficulty from easy to hardall video generators · HumanScore difficulty breakdownHumanScoreHurt
generated videos versus real videosbest video generators · HumanScore leaderboardOverall HumanScore: 94.3 (real) vs 91.1 (best generated)Hurt

Surprising findings

Limitations

Experimental setup

Datasets: Kinetics-700 real videos collected from YouTube, Bilibili, YouTube-8M

Benchmarks: HumanScore

Metrics: HumanScore (0-100 per dimension and overall) Spearman rank correlation with human preference Spearman correlation with VBench axes

Baselines: VBench Wan 2.2 CogVideoX-5B HunyuanVideo 1.5 Kandinsky 5.0 pro Sora-2 Veo 3.1 fast KlingAI 2.5 Turbo Pro Seedance 1.0 Pro fast Hailuo 02

Builds on: VBench HADM PromptHMR MeTRAbs Kinetics-700

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