Other than a microphone, what else would be required to get an Arduino device to recognise a single spoken word with tolerable accuracy...say 85-90% plus? Is this even possible?
I don't know anything about it, but I've heard of the [u]EasyVR Shield[/u].
Accuracy is going to depend on lots of factors and you won't get the "advertised" accuracy unless you are consistently close to the mic in a quiet room. And of course, there are two types of errors... A miss or a false-match.
Thanks. Time to do some research...!
I've seen quite some attempts at integrating Alexa controls with an Arduino. That's another avenue. Arduinos by themselves are not suited for speech recognition (too little memory, and very limited processing power).
If you want to do speech recognition on your own device, look into more powerful systems like the Raspberry Pi.
It’s literally for a parlour trick...if I told you the word I wanted it to recognise, some of the more geeky amongst you would get it immediately. But for the sakes of available space it has to run on teensiest tiniest Arduino I can find (cheapest would be no bad thing too...).
I’m not looking for a Hal 9000 here...and if it gets the word wrong the world will not end...but I’d like to give it a go at least...long term project, so apologies if I don’t post back for a while, but thanks for the input.
Maybe an ESP8266 can pull it off. If you get the QFN package of this chip, and leave out the WiFi antenna they're not much bigger than the tiny.