Towards 3D Human Pose Estimation in the Wild: a Weakly-supervised Approach

ICCV 2017 Xingyi ZhouQixing HuangXiao SunXiangyang XueYichen Wei

In this paper, we study the task of 3D human pose estimation in the wild. This task is challenging due to lack of training data, as existing datasets are either in the wild images with 2D pose or in the lab images with 3D pose... (read more)

PDF Abstract

Evaluation results from the paper


Task Dataset Model Metric name Metric value Global rank Compare
3D Human Pose Estimation Geometric Pose Affordance Baseline model MPJPE (CS) 99.4 # 4
3D Human Pose Estimation Geometric Pose Affordance Baseline model MPJPE (CA) 89.2 # 4
3D Human Pose Estimation Geometric Pose Affordance Baseline model PCK3D (CS) 81.3 # 4
3D Human Pose Estimation Geometric Pose Affordance Baseline model PCK3D (CA) 83.6 # 4
3D Human Pose Estimation Human3.6M Weakly Supervised Transfer Learning Average MPJPE (mm) 64.90 # 19