AI Paper Atlas

Research map / 3D vision / Human pose estimation

PCLs: Geometry-Aware Neural Reconstruction of 3D Pose With Perspective Crop Layers

Frank Yu, Mathieu Salzmann, Pascal Fua, Helge Rhodin

CVPR 2021 · 3D vision › Human pose estimation · approach: Transformer-based multi-view 3D pose · 20 citations (OpenAlex)

Check 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 root centering with PCL and PCL_invMLP 2D-to-3D lifting · H3.6M, 2D GT + 3D Root GTMPJPE (mm): 45.3 \pm 0.2 -> 40.1 \boldsymbol{\pm} 0.5Improved
Replace root centering with PCL and PCL_invMLP 2D-to-3D lifting · MPI-INF-3DHP, 2D GTMPJPE (mm): 74.1 \pm 0.5 -> 50.1 \boldsymbol{\pm} 0.2Improved
Replace rectangular STN cropping with PCLResNet-50 image-to-3D pose model · H3.6M, image inputMPJPE (mm): 96.5 -> 94.1Improved
Replace rectangular STN cropping with PCLResNet-50 image-to-3D pose model · MPI-INF-3DHP, image inputMPJPE (mm): 117.7 -> 109.5Improved
Apply PCL to the temporal pose sequenceT-CNN · H3.6M, original focal lengthMPJPE (mm): 47.3 \pm 0.0 -> 48.8 \pm 0.3Hurt
Use PCL with known focal length instead of unnormalized T-CNN inputT-CNN · H3.6M, changed focal lengthMPJPE (mm): 72.7 \pm 0.5 -> 48.9 \pm 0.2Improved
Reduce MLP + PCL linear size from 1024 to 896 (ablation)MLP + PCL · H3.6M validationMPJPE (mm): 43.8 -> 45.4Hurt
Reduce MLP + PCL linear size from 1024 to 768 (ablation)MLP + PCL · H3.6M validationMPJPE (mm): 43.8 -> 46.5Hurt
Apply only the x-axis PCL rotation post hoc (ablation)MLP + RC · MPI-INF-3DHP, 2D GT + 3D Root GTMPJPE (mm): 69.4 -> 67.6Improved
Apply the x- and y-axis PCL rotations post hoc (ablation)MLP + RC · MPI-INF-3DHP, 2D GT + 3D Root GTMPJPE (mm): 69.4 -> 66.3Improved
Train with PCL rather than applying its rotation only after training (ablation)MLP 2D-to-3D lifting · MPI-INF-3DHP, 2D GT + 3D Root GTMPJPE (mm): 66.3 -> 45.6Improved
Replace rectangular STN cropping with PCLCNN pose reconstruction · Cube Dataset, unseen test setMPJPE (mm): 13.3 -> 8.5Improved

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