Amazon.com, Inc.
Inter-channel level difference based acoustic tap detection

Abstract:

A system configured to detect a tap event on a surface of a device using microphone audio data and motion data. Instead of using a physical sensor to detect the tap event, the device detects a tap event based on a power level difference between two or more microphones. When a power ratio exceeds a threshold, the device may detect a tap event and perform an action, such as delaying or ending an alarm. To reduce false positives caused by wind or loud noises close to the microphones, the device may confirm a tap event using motion data that indicates actual movement of the device. In some examples, the device may detect the tap event using a neural network processing the microphone data and the motion data. In addition, the device may embed the motion data within the microphone data using unused bits of the microphone data.