atlas

Research map / 3D vision / Human pose estimation

2D/3D Pose Estimation and Action Recognition Using Multitask Deep Learning

CVPR 2018 · 3D vision › Human pose estimation · approach: Monocular 2D-to-3D pose lifting · 594 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
Add estimated poses in place of Kinect poses (ablation)Ours on NTU RGB+D · NTU RGB+D (cross-subject)Accuracy (%): +2.9%Helped
Fine-tune the full network end to end (ablation)Ours on NTU RGB+D · NTU RGB+D (cross-subject)Accuracy (%): +3.3%Helped
Average predictions across multiple video clips (ablation)Ours on NTU RGB+D · NTU RGB+D (cross-subject)Accuracy (%): +1.1%Helped
Train 3D pose estimation on Human3.6M + MPII instead of Human3.6M only (ablation)Ours · Human3.6MMPJPE (mm): 67.3 -> 55.1Helped
Use multi-crop evaluation with horizontal flips instead of single-crop evaluation (ablation)Ours · Human3.6MMPJPE (mm): 55.1 -> 53.2Helped
Use annotated poses instead of estimated poses (ablation)Ours on Penn Action · Penn ActionAccuracy (%): 97.4 -> 98.6Helped
Train pose and action jointly rather than sequentially followed by fine-tuning (ablation)Ours on Penn Action · Penn ActionAccuracy (%): 97.4 -> 97.3; 0.1%Hurt

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