You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The continuous param doesn't work properly in Chrome desktop or for Android. It stops when no-speech was detected after usually 3-4 seconds. I think this is done to reduce the traffic, since they transfer the audio to the Google servers to analyse them.
Some implementations just restart the recognition by calling the start method in the onend event as a workaround. This works for desktop, but on Android devices you'll will always hear a connect/disconnect sound, which is rather annoying.
It would be nice to enable the microphone continuously, but (re)start the main processing only when certain keywords/phrase have been detected locally without network connection, like the 'Hey Google' phrase for the voice assistant. Something like:
constrecognition=newSpeechRecognition();// continuously enable microphone to listen to the userrecognition.continuous=true;// phrase to start the main processingrecognition.startphrase='Hey Web';recognition.start();recognition.onprocessstart=function(){console.log('main processing has started');}recognition.onprocessend=function(){console.log('main processing has ended');}// will only be called when startphrase has been saidrecognition.onresult=function(){ ... }
The text was updated successfully, but these errors were encountered:
The
continuous
param doesn't work properly in Chrome desktop or for Android. It stops when no-speech was detected after usually 3-4 seconds. I think this is done to reduce the traffic, since they transfer the audio to the Google servers to analyse them.Some implementations just restart the recognition by calling the
start
method in theonend
event as a workaround. This works for desktop, but on Android devices you'll will always hear a connect/disconnect sound, which is rather annoying.It would be nice to enable the microphone continuously, but (re)start the main processing only when certain keywords/phrase have been detected locally without network connection, like the 'Hey Google' phrase for the voice assistant. Something like:
The text was updated successfully, but these errors were encountered: