This demo shows you how to use voice-recognition API. This only works with Chrome on a laptop.
We setup a new instace of webkitSpeechRecognition and save a references to the variable recognition for later use.
The instance supports a couple of events like start, end, result that we can use to hook up functions.
onClick will make the isntance listen for input through the computer microphone.
You will be prompted to allow this, due to Chromes built-in security layer.
onStart will fire when you allowed access to the microphone and started talking :-)
onEnd will fire when you stopped talking.
onResult will fire when there is match from the voice-recognition service.
It provides an argument e where we can find various information such as the recognized text.
All commands are located in the commands variable and are case sensitive.
Things to try:
- Add a new command and function of your own
- Can you make it so the user doesn't have to say the exact phrase? Can it be more forgiving?
Read more: