I was wondering if anyone knew how Vuforia's model detection and CPU/GPU usage was affected by a Model Target's poly count.
For example let's say we have a 100K poly source model used to generate a model target and model detection works well on device.
If we bump up the source model to 200K and generate a new model target would it be expected that model detection would work better, but Vuforia detection processing would increase?
I'm trying to figure out the trade offs and effect a source model's poly count would have on the runtime detection and performance.
Ideally we could reach a balance point where detection is good enough and runtime performance is also as good as it can be.
Right now our app generates a lot of heat and drains our test mobile devices very quickly so we're always looking for ways to reduce power consumption.