search

google-deepmind / Temporal-3D-Pose-Kinetics

Exploiting temporal context for 3D human pose estimation in the wild: 3D poses for the Kinetics dataset
http://arxiv.org/abs/1905.04266
Apache License 2.0
219 stars 27 forks source link

2D keypoints used for fitting #4

Closed mkocabas closed 5 years ago

mkocabas commented 5 years ago

Hi @hmph and @cdoersch,

Thanks for releasing the dataset. Is it also possible to release the 2D keypoints used for fitting, too? Thanks!

cdoersch commented 5 years ago

Greetings @mkocabas , the 2d keypoints are already available in the dataset: see the '2d_keypoints' field.

mkocabas commented 5 years ago

Thanks for your reply. 2d_keypoints field contains the projected keypoints using SMPL. However, I intend to use the detected 2D keypoints obtained through Papandreou et al.'s method [1].

[1] G. Papandreou, T. Zhu, N. Kanazawa, A. Toshev, J. Tompson, C. Bregler, and K. Murphy. Towards accurate multi-person pose estimation in the wild. In CVPR, 2017

cdoersch commented 5 years ago

Sorry, we don't currently provide those, nor do we have any plans to.

mkocabas commented 5 years ago

Thanks for the reply!