I have an existing Keras model that I wish to deploy on my drone, using the GPU. (Dense depth estimation from a single camera.)
The drone has a Qualcomm chip set - ARM 820, Adreno GPU - not an ARM GPU. I've followed the examples for ARM CL and have tested that one of the examples will work, using OpenCL. Great.
However, the path seems a little steep to deploy my own model. ARM NN seems to show a better path, but it's unclear from the documentation if it supports anything other than ARM GPUs.
Looking for some ideas/advice on the best way forward. (Or heck, should I be trying out tvm or another approach?)
This chart is from https://developer.arm.com/ip-products/processors/machine-learning/arm-nn
ARM NN can support other third-party IP such as GPU. The prerequisite is that the third-party IP can provide the necessary drivers and SW functions.
View all questions in Machine Learning forum