search

hongsukchoi / 3DCrowdNet_RELEASE

Official Pytorch implementation of "Learning to Estimate Robust 3D Human Mesh from In-the-Wild Crowded Scenes", CVPR 2022
MIT License
155 stars 15 forks source link

Have question about the conversion between cam_param and cam_trans #24

Open JosephKKim opened 1 year ago

JosephKKim commented 1 year ago

https://github.com/hongsukchoi/3DCrowdNet_RELEASE/blob/fde86df0fb8799873cd19c683b8251b16640acb3/main/model.py#L38

Hello @hongsukchoi First, thanks for the nice work you've publicly released! Currently I'm doing a research on the multi-person reconstruction and while reading your code line by line I encountered the function get_camera_trans which I mentioned above.

As far as I understand, from RotationNet we get cam_param and the function converts it into cam_trans (which stands for camera translation matrix) and again it goes in to get_coord function.

My QUESTION starts here! I don't understand the exact meaning or assumption of gamma and k_value. Is there any theory referring to it?

You seems like replying to issues so.. Thanks in advance!

Joseph

hongsukchoi commented 1 year ago

Hi!

You can refer to this paper and code; https://github.com/mks0601/3DMPPE_POSENET_RELEASE https://arxiv.org/pdf/1907.11346.pdf

If you are Korean, you can also look on this: https://redstarhong.tistory.com/97

JosephKKim commented 1 year ago

Oh! okay thanks for the fast response! I know the paper 3DMPPE but couldn't connect it with the code. So in this code RotationNet has the role similar to (but inferencing more parameters than RootNet with its multi-branch) the RootNet in 3DMPPE is that right?

Thanks again for the fast response. It helped a lot. Best, Joseph.

hongsukchoi commented 1 year ago

Yes, you are right.

As you may know, k value is the scale factor obtained in a closed equitation, and gamma is correcting it in a data-driven way.