Anomaly detection for voice controlled devices

A computer-implemented method comprising:

• receiving a multi-dimensional input associated with audio data representing an utterance received by a speech-detection device, the multi-dimensional input comprising a speaker identifier (ID) of a speaker of the utterance and a device ID of the speech-detection device;
• processing the multi-dimensional input using a first machine learning model associated with at least one of the device ID or the speaker ID to determine a first anomaly rating for the utterance;
• determining that the first anomaly rating is above a first anomaly threshold;
• determining that the utterance is classified as an anomaly;
• and
• disabling one or more features for the speech-detection device.

A system comprising:

• at least one processor;
• and
• at least one memory including instructions that, when executed by the at least one processor, cause the at least one processor to:
• receive a multi-dimensional input associated with audio data representing an utterance received by a speech-detection device, the multi-dimensional input comprising a speaker identifier (ID) of a speaker of the utterance and a device ID of the speech-detection device;
• process the multi-dimensional input using a first machine learning model associated with at least one of the device ID or the speaker ID to determine an anomaly classification for the utterance;
• determine that the utterance is classified as an anomaly;
• and
• disable one or more features for the speech-detection device.