Robust Speech Recognition?

[surovich]

Thought I would ask before I go to write something ....

Is there anything in Homeseer now that will allow a more robust Speech Recognition?

I cant be limited to entering every term in the speech recognition for an event - I have a weather routine that will allow you to ask "What was the weather like in 1982" (Or any year for the last 100 years)

That's just one example ....

Thanks
Scott

[Rupp]

You may want to write your own as you can put wildcards in for events but it may take longer that writing your own.

[surovich]

Thanks for the reply ....

So I can basically use "What was the weather in *" and pull the *, does HS have a Speech to text option?

[surovich]

Ok, I've got my routine written in Python - I can capture any sentence and convert it to text - So I'm covered there. Now I think my issue will be picking a recording device since the input could come from any speaker client and my routine by default will pick the default sound input....

Can Homeseer record a wav file from a VBS script? (Since homeseer knows the speaker client in use)