Voice Recognition

Questions and comments specific to a particular plugin should go here.
Post Reply
wacind
Posts: 20
Joined: Fri Jan 30, 2015 6:50 pm

Voice Recognition

Post by wacind » Wed Feb 25, 2015 4:08 pm

Hi

I updated an older voice recognition plugin (speakeasy - http://www.eventghost.org/forum/viewtop ... =10&t=6024)

Ive updated it with extra logging to indicate the confidence level of the words or phrases used, added two sets of phrase confidence levels and added a filler words section to add words or phrases to help eliminate false positives.

More information on how to use it and tips to improve recognition can be found at https://github.com/wacind/VoiceRecognition

please keep in mind that this is a work in progress and works fine for me. let me know if you have any issues or need help setting up an eventghost tree with this

SyntaxSemantic
Posts: 6
Joined: Thu Jan 15, 2015 4:13 am

Re: Voice Recognition

Post by SyntaxSemantic » Tue Feb 09, 2016 2:45 am

I tried the earlier version of SpeakEasy and did not have much luck. My biggest problem was that everytime I enabled the plugin, and set up the keywords, the event would fire itself without any spoken keywords. Is this something expected or it just an isolated incident. Any suggestions would be appreciated.

wacind
Posts: 20
Joined: Fri Jan 30, 2015 6:50 pm

Re: Voice Recognition

Post by wacind » Tue Mar 08, 2016 9:11 pm

So the best way i found to configure this would be to use phrases instead of keywords. For example "roger start music player,roger start notepad" instead of the individual keywords. This reduces false positives

SyntaxSemantic
Posts: 6
Joined: Thu Jan 15, 2015 4:13 am

Re: Voice Recognition

Post by SyntaxSemantic » Wed Mar 16, 2016 3:56 am

I found a better version of the plugin through GitHub that you had modified the Py scripts allowing the user to specify accuracy values for the spoken words and phrases. Works MUCH better now!!! :D The one that I was having problems with was the original GitHub post from maikhorma

Here's the link if anyone is interested :
https://github.com/wacind/VoiceRecognition

Thank you wacind!!! :wink:

Post Reply