How to approach to parallelization of algorithms?
I want to parallelize algorithm (object tracker). How should I approach to this topic? Is this only changing namespace and syntax to gpu or is this more demanding activity?
have you got any solution? i have to do the same on jetson tk1