Voice/motion control

MichMich

coolbotic

Hope it works now haha !

Simon

Hi
i haven’t put voice control into my mirror but have played around with it for a kitchen based recipe generator. (Say Baked Alaska and the display pulled a recipe off BBC food website and displayed it. )

Basically voice recognition is fairly simple. What is difficult is getting it to start listening.
The easiest way of doing it is to use the google api Here is a good source. http://stevenhickson.blogspot.co.uk/2013/06/voice-command-v30-for-raspberry-pi.html

The problem I had when I played with this is that. It records what you say and sends it to google for recognition. If you want a system that is always listening then you would end up sending recordings to google 24/7. . You could try using http://cmusphinx.sourceforge.net/ to recognize an initial key word then turn on google recognition. But I couldn’t really get a great response from that. However I have a strong Irish accent which doesn’t help :)
So you might have more luck. Also a good microphone will make your life easier.

MichMich

@Simon Hi Simon. Do you know if the google API is supported under Electron?

PtrBld

It seems that Google has removed this feature for electron, however when you use chromium together with MM² it may works… See this stackoverflow question.

Simon

Don’t know. I coded with it in python. To tell you the truth I have only heard of Electron about 10 minutes ago :) My magic mirror is coded in python (far more hacky and inelegant then your solution) but it entertained me for a while. :)

paviro

electron-speech, alexa-voice-service & pocketsphinx might be worth a look :)

coolbotic

@Simon Have you posted you code anywhere would love to see it!!!

Simon

@coolbotic No it’s not very pretty :) I’ll see what I can do

matt08

What do you think about using wit.ai to voice control MM?
Free, unlimited API that could be implemented