I noticed the setting for human height in the SV3DT config. Does it use this to aid resiliency of tracking or just to figure out the foot location?
Lets say you set it to some average-ish human height like 5’ 8" (1.72m). Is the result just going to be that taller people are going to have a predicted foot location closer to the camera than reality and shorter people are going to have a predicted foot location further from the camera than reality?
How would it do in the extremes like a child or someone 6’ 4".
How does it perform when people go to sit like on a bench.
Thanks
**• Hardware Platform Jetson
**• DeepStream Version 7.0