Smart speaker device with cognitive sound analysis and response

摘要

Smart speaker mechanisms are provided for processing audio sample data captured by the audio capture device is provided. The mechanisms capture one or more audio samples from a monitored environment and store the captured one or more audio samples in the audio sample buffer. The one or more captured audio samples are analyzed to determine whether the one or more captured audio samples represent a potential wake sound for initiating a responsive action by the smart speaker device. The potential wake sound is a variable sound determined to not be a normal ambient sound of the monitored environment. In response to determining that the one or more captured audio samples represent a potential wake sound for initiating a responsive action by the smart speaker device, the mechanisms initiate performance of the responsive action by the smart speaker device.