As in the title, I have significant experience in embedded development. I’m searching for some guidance on moving my career toward Edge AI but I can’t find learning materials or case studies I could use on my own to put them in my portfolio.
I wish to find anything valuable so I can:
Learn some concepts in a way it should be done step by step.
Further extend that knowledge practicing on my own.
I wish to find how to work with ML/AI on microcontrollers and embedded linux (on my Raspberry Pi) but also to apply it into something practical. Let’s assume that working on Arduino environment is too basic but I will also appreciate it.
I can’t provide very much guidance here, but when you have very low computing horsepower available, you will have limitations on the size of problems you can work on.
For example, you can either run very simple models very quickly, or you can run complex models very slowly. But you can’t run deep high complexity models in real time.
The reason you can’t find much material to teach this online is because 1) the field is rather new, and 2) experience is much more valuable than tutorials. It takes years to accumulate the experience required to make good decisions among the many tradeoffs you may have to assess.
Thanks for your answer. Sorry for no response for a while.
Yes, I understand. But I feel like playing around with my Raspberry PI won’t get me anywhere I think that doing AI on embedded offers much more than porting models to smaller devices.
I’m starting my experiments with CMSIS-NN but I wish to find a learning guide for that…
I think that doing AI on embedded offers much more than porting models to smaller devices.