-3

I have searched online and all implementations seem to use Raspbian or Linux. I would like to do be able to do voice control on my raspberry pi baremetal. Any suggestions?

adecyber
  • 1
  • 1
  • of course it is possible, anything you can do on an os you can do without an os, its just a processor. pi docs are tolerable. but there is a very strong baremetal community for the pi that doesn exist at this level for other platforms. you "simply" need to implement whatever system calls this software wants. note: just run it on linux, maybe have a second pi or processor for your baremetal work...its not going to be worth the effort unless this is a climb the mountain "because it is there" kind of thing. " – old_timer Mar 21 '18 at 14:41

1 Answers1

0
  1. Raspberry Pi is terrible for bare metal. Broadcom isn't exactly known for being forthcoming about the low level details of the RPi.

  2. Voice recognition has a huge huge overhead. Even if you wrote a voice recognition program in assembly (which would probably take longer than your natural life), the RAM requirements would still far exceed the capabilities of the RPi. That doesn't get into the requirements of the internal logic of Alexa/Siri/Cortana, which I imagine are similarly astronomical.

  3. If you are truly hellbent on trying, you pretty much have to build everything up from scratch if you really want to go bare metal. Can you write all of the device drivers?

  4. The real problem is that even if you write all of the device drivers, I doubt Amazon is going to give you access to the source so you can cross compile it. Plus you'd have to write an entire library from scratch.

So long story short, no you can't sadly.