0

I am trying to make an offline voice controlled home assistant. I use pocketsphinx for speech to text and espeak for text to speech, like if i give a command "open browser", then it opens the browser and speaks back "open browser processed".

The problem is when I connect to an external speaker, pocketsphinx recognises the "open browser processed" and this goes into a loop.

How can I configure pocketsphinx to pause for a little more time before recognising the next word?

I am using the livespeech function and code with Python.

Nikolay Shmyrev
  • 24,897
  • 5
  • 43
  • 87
Pratik Kumar
  • 2,211
  • 1
  • 17
  • 41
  • Simply do not start recognition while TTS playing. Related questions http://stackoverflow.com/questions/39625265/android-speech-recognition-conflicting-with-text-to-speech http://stackoverflow.com/questions/34816491/how-to-stop-speech-recognition-engine-listening-its-own-speech http://stackoverflow.com/questions/16400840/stopping-speech-recognition-before-using-text-to-speech – Nikolay Shmyrev Mar 17 '17 at 14:50
  • @NikolayShmyrev Can the livespeech function be configured, such that after recognising a word, it stops for 3 seconds then again starts recognising? – Pratik Kumar Mar 17 '17 at 19:58
  • You can simply use blocking tts inside speech loop, it does not start recognition until you inside the for loop. – Nikolay Shmyrev Mar 17 '17 at 20:08

0 Answers0