Instructions for Combining Datasets (adding a foot model)

[ConnorM17]

Thanks for all the great work you’ve done! I have read some other people issues regarding combining the OpenPose foot dataset with the current Hyperpose models. From what I have read it is possible to do this, however, there doesn’t seem to be a comprehensive explanation on how to do this. It would greatly increase the usefulness of the current Hyperpose models for things like movement analysis.

I was wondering if one of the main contributors could provide a thorough explanation on how to do this? I am particularly interested in the addition of the keypoints seen here (https://cmu-perceptual-computing-lab.github.io/foot_keypoint_dataset/). Or, if there are any other people that have successfully added the foot keypoints I would really appreciate some help.

Thanks!

[Gyx-One]

Hello @ConnorM17 !
Pelease wait for a few days, Pifpaf is under debugging and will be finished soon! I'll add Hand and Foot dataset and instruction over cobining dataset then. I've been busy these days and it will take a few time. :)

[ConnorM17]

Thank you! Yes, please that would be great. Looking forward to getting the food and hand added!

[ConnorM17]

Hey @Gyx-One ! Any updates on this?

[ConnorM17]

Has anyone else had any success doing this?