Wed-1-2-3 An Audio-Based Wakeword-Independent Verification System

Abstract: We propose an audio-based wakeword-independent verification model to determine whether a wakeword spotting model correctly woke and should respond or incorrectly woke and should not respond. Our proposed model works on any wakeword-initiated audio by operating only on the audio surrounding the wakeword, allowing the model to function on custom, personalized, and unknown wakewords. This model is based on two key assumptions: that audio surrounding the wakeword is informative to determine if the user intended to wake the device and that this audio is independent of the wakeword itself. We show experimentally that on new wakewords not included in the training set, our model trained without training examples or knowledge of the wakeword is able to achieve verification performance comparable to models trained on 5,000 to 10,000 annotated examples of the new wakeword.