Accuracy: 93
Papers with Code · CC BY-SA 4.0
Pose Estimation
Pose Estimation is a computer vision task where the goal is to detect the position and orientation of a person or an object. Usually, this is done by predicting the location of specific keypoints like hands, head, elbows, etc. in case of Human Pose Estimation.
A common benchmark for this task is MPII Human Pose
<span style="color:grey; opacity: 0.6">( Image credit: Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose )</span>
논문
Average Precision: 0.357Average Recall: 0.445
PCKh@0.5: 88.6
Mean Reconstruction Error (mm): 13.5
A2J: Anchor-to-Joint Regression Network for 3D Articulated Pose Estimation from a Single Depth Image
Mean mAP: 88.0
PCKh-0.5: 81.1
Mean mAP: 88.75
PCKh@0.5: 91.2PCKh@0.1: 36.0
Mean Reconstruction Error (mm): 12.2
Average End-Point Error: 10.81
Mean Reconstruction Error (mm): 9.2
Mean Reconstruction Error (mm): 15.2
mAP: 56.5
PCKh-0.5: 80.9
PCKh@0.5: 93
PCKh@0.5: 92.1
Percentage correct: 74.55
Mean Reconstruction Error (mm): 14.3
PCKh@0.2: 92.3
PCKh@0.5: 88.8PCKh@0.1: 22.5
mAP: 56.9
Mean mAP: 93.38
Mean mAP: 89.19
Mean mAP: 84.9
Mean mAP: 77.4
Mean mAP: 88.74