AI Paper Atlas

Research map / 3D vision / Human pose estimation

OmniPoser: Flexible Human Motion Recovery in the Wild with Masked Flow Matching

Minghao Liu, Tutian Tang

ECCV 2026 · 3D vision › Human pose estimation · approach: Physics and sensor-based motion capture · 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
replace learnable per-joint embeddings (MCE) with zero-padding (ablation)OmniPoser · AMASS upper-bodyMPJPE (mm): 88.1 -> 32.1Improved
use OmniPoser Ours-S instead of RoHMsparse observation · Nymeria upper-bodyMPJPE (mm): 59.3 -> 21.1Improved
use OmniPoser Ours-S instead of RoHMsparse observation · AMASS upper-bodyMPJPE (mm): 34.8 -> 32.1Improved
use OmniPoser instead of RoHMsparse observation · AMASS upper-bodyacceleration error: 2.3 -> 2.7Hurt
train unified Ours-U instead of setting-specific Ours-SOmniPoser · AMASS random joint maskingMPJPE (mm): 35.7 -> 49.3Hurt
use OmniPoser decoupled global/local streams instead of EgoEgo coupled predictionhead 6D pose · AMASSMPJPE (mm): 167.4 -> 83.2Improved
use OmniPoser instead of EgoAllohead 6D pose · AMASShead translation error (mm): 6.2 -> 17.1Hurt
use OmniPoser instead of SIP6 IMUs + keypoints · TotalCapturetranslation error (mm): 5.42 -> 15.9Hurt
use OmniPoser decoupled generative architecture instead of RobustCap regression6 IMUs + keypoints · AIST++jerk: 2.49 -> 0.29Improved
increase Euler ODE steps beyond 1OmniPoser · benchmarksaccuracyNo clear effect
increase ODE steps (diversity)OmniPoser · diversity analysissample diversity: 1 step -> 5 steps @250 fps, 30 steps @80 fpsImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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