I've had a good browse around the topics in this group trying to see if this question has been addressed before, but couldn't find exact thing. So, apologies if this is old hat. Please also excuse the newbie nature of my question.
How can I have an iOS app recognise a phrase from voice input only if the intensity hits certain decibel? Ex:When the user says "hello there", it should capture the voice and then check the phrase to make sure whether user said hello there or something else. Perhaps, openears might solve this for me. But how can I measure the intensity of voice? Voice input must be considered only if the phrase "hello there" is spoken at 110dB. If not, it should ignore. Can this be done with Openears? Please guide me with your suggestions
OpenEars developer here. OpenEars' PocketsphinxController class (the class that does speech recognition) has a property pocketsphinxInputLevel that you can read on a secondary thread to find out the input power. There's info about it in the docs and a copy-paste-able example of its usage in the OpenEars sample app so that should get you started. The decibel levels use the same convention as the AudioQueue decibel readings so they give decibels as a negative offset from zero where zero is the clipping point, instead of a positive absolute value.