AI Paper Atlas

Research map / 3D vision / Human pose estimation

Learning Human Mesh Recovery in 3D Scenes

Zehong Shen, Zhi Cen, Sida Peng, Qing Shuai, Hujun Bao, Xiaowei Zhou

CVPR 2023 · 3D vision › Human pose estimation · approach: Human mesh recovery from images · 16 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
predict scene-aware root position and mesh in a single forward passMETRO · RICHG-MPJPE: 511.7 -> 264.6Improved
predict scene-aware root position and mesh in a single forward passMETRO · PROXG-MPJPE: 265.6 -> 150.4Improved
replace scene-aware optimization with a learned single-pass methodMETRO + SA-Opt · RICHG-MPJPE: 563.1 -> 264.6Improved
add scene-aware root and contact estimation (ablation)SA-HMR root module · RICHRtErr: 510.8 -> 284.7Improved
add scene-aware root and contact estimation (ablation)SA-HMR root module · PROXRtErr: 364.2 -> 132.3Improved
add scene-aware mesh recovery after root refinement (ablation)SA-HMR · RICHRtErr: 284.7 -> 246.5Improved
add scene-aware mesh recovery after root refinement (ablation)SA-HMR · PROXRtErr: 132.3 -> 111.8Improved
use a parallel cross-attention scene network instead of early feature fusion (ablation)SA-HMR · RICHG-MPJPE: 304.8 -> 264.6Improved
use a parallel cross-attention scene network instead of early feature fusion (ablation)SA-HMR · RICHCErr: 10.2 -> 8.9Improved
replace estimated root with ground-truth root (ablation)SA-HMR mesh recovery · RICHMPJPE: 93.9 -> 89.2Improved
replace estimated contacts with ground-truth contacts (ablation)SA-HMR mesh recovery · RICHMPJPE: 93.9 -> 90.4Improved
replace estimated root and contacts with ground truth (ablation)SA-HMR mesh recovery · RICHMPJPE: 93.9 -> 76.7Improved

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