Hello,
I am currently working on a pose estimation project (with a ZED 2i), where the objective is to make use of joint angles to assess athletes’ performance in sports.
I have been using the ZED SDK and was interested in the body tracking module, which seems to rely on a pose detection model that integrates depth data.
However, I was not able to find specific details regarding the used model, information I wish to include in my literature review of available methods :
- Does the model have a name, or is it based on a known architecture?
- Was it developed in-house, if so, is there any publication or paper describing it?
- Was any public dataset used for training?
I am purely asking for academic purposes, if certain aspects are confidential, I completely understand. If any technical references or details are publicly available, please point me to them.
Thank you in advance for your help.