Hotword Detection

Recognise spoken words from 1 second audio clips. Use the tensorflow speech_commands dataset to train a model to recognise spoken words from a fixed vocabulary. Then use this model to predict words spoken by the user into the computer's microphone.