Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

how the 3d model runing? #2

Open
gebibaiyang opened this issue Jan 2, 2019 · 3 comments
Open

how the 3d model runing? #2

gebibaiyang opened this issue Jan 2, 2019 · 3 comments

Comments

@gebibaiyang
Copy link

Hi, I am very happy when saw this project, because I have the similar demand。so I want to know the fundamental. There are some problems:

  1. what is 3d_data0*.text ?
  2. what is the operations when the 3d_data0 to 3d animation.

I am waiting for your replay. Thank you very much.

@keel-210
Copy link
Owner

Sorry. I didn't notice this issue,the github notification wasn't working.
1.3d_data * .text is attitude information in 3d space calculated by tf-openpose.
2.I am not good at English, so I can not understand the meaning of Question 2 completely, but I will answer for the time being. 3d_data * .text to animation conversion tried both IK and FK. However, the basics are the same in both cases, and the indirect vector from the connected joint is normalized and used as posture information. Note that the posture information in the text is z-up, but unity is y-up.
I am sorry for more information in Japanese, please refer to the following link.
tf-openposeによる姿勢推定をunityに落としこもうとしたけど失敗した話
In addition, I supplement the detailed information in the first issue.

@timtensor
Copy link

Hi thanks , for the information. I dont want anything real time . I just want to input a video , get the 3d pose estimation following this link which basically life 2d joints to a 3d space

https://github.com/ArashHosseini/3d-pose-baseline

If i understand it correctly , i think somehow the entire 3D c0-ordinates have to be written into perhaps a txt file , the txt file is then translated into unity 3d model . Have you had any luck on that ?

@keel-210
Copy link
Owner

keel-210 commented May 9, 2019

Hi.
It seems that I was misunderstanding.
Your interpretation is correct. This Unity project reflects tf-openpose's attitude estimation results on Unity's Humanoid 3d model. All 3d pose estimation is done by tf-openpose, and this project can not input video.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants