search

akanazawa / hmr

Project page for End-to-end Recovery of Human Shape and Pose
Other
1.55k stars 395 forks source link

How to use 3d key points to fit the human model #61

Closed zhanghr123 closed 5 years ago

zhanghr123 commented 5 years ago

Hello, I am very interested in your project. I found that fitting a human body shape and posture from a single picture works well. But the effect is not very good when applied to the video. Your recent work, Learning 3D Human Dynamics from Video, I am also learning.I recently saw some work, such as https://github.com/facebookresearch/VideoPose3D. I would like to ask you how to apply 3D key points to the smpl model.I look forward to your reply.

akanazawa commented 5 years ago

You could always try to fit SMPL to 3D joints (take the approach of SMPLify, but think of 3D joint input instead of 2D), but I have never tried this. The video fitting code for our siggraph paper is coming, as well as the recent paper hopefully soon, so stay tuned! Thanks!

davidpagnon commented 4 years ago

I'm also interested, if anyone knows of any approach to this end!