Method and device for audio input routing

A device comprising:

• at least one processor;
• and
• memory storing instructions that, when executed, cause the at least one processor of the device to:
• monitor an audio input for an occurrence of a trigger phrase that is spoken by a user of the device while the device is in a low power mode;
• and
• in response to detecting the occurrence of the trigger phrase:
• wake the device from the low power mode;
• and
• buffer the audio input while the device wakes up from the low power mode;
• and
• in response to the device waking up from the low power mode, process the buffered audio input to detect an occurrence of a command phrase that is spoken by the user in a portion of the buffered audio input;
• in response to detecting the occurrence of the command phrase that is spoken by the user in the portion of the buffered audio input, determine whether the command phrase spoken by the user of the device is directed to the device or is located in the middle of a conversation, wherein the instructions to determine whether the command phrase spoken by the user of the device is directed to the device or is located in the middle of a conversation comprise instructions to:
• determine whether a length of the buffered audio input satisfies a threshold length;
• determine, based on the length of the buffered audio input satisfying the threshold length, that the command phrase spoken by the user is located in the middle of the conversation;
• and
• in response to determining the command phrase spoken by the user is located in the middle of the conversation and based on the length of the buffered audio input satisfying the threshold length, invalidate the command phrase.

A non-transitory computer-readable storage medium storing instructions that, when executed, cause at least one processor of a device to perform operations, the operations comprising:

• monitoring an audio input for an occurrence of a trigger phrase that is spoken by a user of the device while the device is in a low power mode;
• and
• in response to detecting the occurrence of the trigger phrase:
• waking the device from the low power mode;
• and
• buffering the audio input while the device wakes up from the low power mode;
• and
• in response to the device waking up from the low power mode, processing the buffered audio input to detect an occurrence of a command phrase that is spoken by the user in a portion of the buffered audio input;
• in response to detecting the occurrence of the command phrase that is spoken by the user in the portion of the buffered audio input, determining whether the command phrase spoken by the user of the device is directed to the device or is located in the middle of a conversation, wherein determining whether the command phrase spoken by the user of the device is directed to the device or is located in the middle of a conversation comprises:
• determining whether a length of the buffered audio input satisfies a threshold length;
• determining, based on the length of the buffered audio input satisfying the threshold length, that the command phrase spoken by the user is located in the middle of the conversation;
• and
• in response to determining the command phrase spoken by the user is located in the middle of the conversation and based on the length of the buffered audio input satisfying the threshold length, invalidating the command phrase.