Virtual Machine, New contributor

[Nanubala Gnana Sai]

Hi,

I'm NG Sai from Indian Institute of Information technology, Sri City. Machine learning has always struck a chord in me, but more than that, I've found a niche in ML on Edge. The idea of running complex models on devices of Kilobytes of memory is revolutionary in my opinion.

My expertise lies with modern c++ and I'm well familiar with working on open source projects.

My main concern was that I can't afford the devices that were listed on the supported devices. Can the build work on Arduino Uno, or does any of the devices mentioned have virtual machine?

Looking forward for a reply!

Good day!

[Jonathan Torkelson]

Hi Nanubala,

I am currently in the process of trying to get TFLite up and running in Visual Studio, using Virtuoso (www.virtuoso-software.com). This would allow TensorFlow Lite to run in a completely virtual environment, running inside Visual Studio with all of the great trappings of the Visual Studio IDE. The Virtuoso virtual environment is a low-code environment that allows things like touch screens, simulated I/O and even real-world I/O effortlessly integrated to interface to the virtual microcontroller. 

I'd suggest watching these videos to see how easy it is to pull in I/O from Click boards (https://www.mikroe.com/click):

https://youtu.be/PpOC7-fB-gw

https://youtu.be/PI-XqjKQZW8

The beauty of this workflow is that it is CHEAP and effortless to setup. Just download an example project, and you are up and running, performing inference and stepping through C/C++ code in Visual Studio. Definitely hoping to get people like you involved. 

Cheers!

Jonathan