I'm developing a robot to help elder persons and I need a little help.
To simplify the process and the hardware, I'm thinking in a first step use oak-d/oak-d-lite to detect/locate/identify the person and then estimate the pose of this person - to check if the older is in a dangerous.
I could detect persons without problem with the examples, but can I run the pose estimation on the same pipeline? there is a way to run 2 pipelines- one for person detection/location and a second one to estimate the pose ?
Thanks a lot.