atlas

Research map / 3D vision / Human pose estimation

Deductive Learning for Weakly-Supervised 3D Human Pose Estimation via Uncalibrated Cameras

AAAI 2021 · 3D vision › Human pose estimation · approach: Monocular 2D-to-3D pose lifting · 15 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
use DWSL instead of Wang, Kong, and Lucey (2019)weakly supervised 3D pose estimation · Human3.6MMPJPE: 83.0 -> 76.7Helped
train DWSL on H36M + MPI rather than H36M aloneDWSL · MPI-INF-3DHP test setPCK: 86.3 -> 88.1Helped
train DWSL on H36M + MPI rather than H36M aloneDWSL · MPI-INF-3DHP test setAUC: 49.8 -> 50.5Helped
add Ski-PosePTZ multi-view training dataDWSL trained on H36M-MV · Ski-PosePTZ test setPMPJPE: 108.7 -> 74.7Helped
remove positive depth rule (ablation)DWSL · Human3.6MPMPJPE: 58.6 -> 73.2; 14.6Hurt
remove symmetry length rule (ablation)DWSL · Human3.6MPMPJPEHurt
remove valid angle rule (ablation)DWSL · Human3.6MPMPJPE: 58.6 -> 104.0; 45.4Hurt
remove data augmentation (ablation)DWSL · Human3.6MPMPJPE: 58.6 -> 62.1; 3.5Hurt
remove reconstruction loss (ablation)DWSL · Human3.6MPMPJPEHurt

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