Setting a maximum listening time about pocketsphinx HOT 7 CLOSED

You can easily do it yourself.

from pocketsphinx.

That is what I did.

from pocketsphinx.

Not sure what the problem is then. Feel free to elaborate.

from pocketsphinx.

I wanted a flag to be passed in that would cut off the recognizer to flush the results after a certain amount of time. The TV is at the perfect tone that it won't stop running until the show stops, I am using the library for home automation and all of my commands are about 5 seconds in length, so I didn't need it to wait for long dictation. I was curious if this was already in the library, but instead I added it myself.

from pocketsphinx.

Something I noticed in pocketsphinx is that it would recognize speech input from ambient sources such as TV/music/background noises without actually changing the hypothesis.

My pseudocode looks like:

if hypothesis !changed -> call onResult;

Now my app works pretty good in noisy environments. Not sure if this is the prefered way.

from pocketsphinx.

@sdejonckheere , I'm sorry, I don't understand your proposal either. For continuous listening we recommend keyword spotting model.

from pocketsphinx.

This is a topical issue for me, @mingram8 how you do it?

from pocketsphinx.