Carnegie Mellon University
Existing human pose datasets contain limited body part types. The MPII dataset annotates ankles, knees, hips, shoulders, elbows, wrists, necks, torsos, and head tops, while COCO also includes some facial keypoints. For both of these datasets, foot annotations are limited to ankle position only. However, graphics applications such as avatar retargeting or 3D human shape reconstruction require foot keypoints such as big toe and heel. Without foot information, these approaches suffer from problems such as the candy wrapper effect, floor penetration, and foot skate. To address these issues, a small subset of foot instances out of the COCO dataset is labeled using the Clickworker platform. It is split up with 14K annotations from the COCO training set and 545 from the validation set. A total of 6 foot keypoints are labeled. We consider the 3D coordinate of the foot keypoints rather than the surface position. For instance, for the exact toe positions, we label the area between the connection of the nail and skin, and also take depth into consideration by labeling the center of the toe rather than the surface.
Foot Keypoint Annotations (Training: ~13.5k annotations, Validation: ~0.5k annotations)
|Download the train2017_foot_v1.zip JSON zip file.|
|Download the val2017_foot_v1.zip JSON zip file.|
|Download the image dataset from the original body-only dataset website (in particular 2017 Train images [118K/18GB] and 2017 Val images [5K/1GB])|
|Trained foot detector models and testing code released in OpenPose|
For training code, check https://github.com/CMU-Perceptual-Computing-Lab/openpose_train.
The software and Foot Keypoint Annotations entitled train2017_foot_v1.zip and val2017_foot_v1.zip, as provided on this webpage, are owned by Carnegie Mellon University and are licensed under the Creative Commons Attribution 4.0 License. All rights not specifically granted are reserved. For any inquiries, please contact Scott McEvoy at email@example.com or firstname.lastname@example.org.