Preventing inadvertent wake in a speech-controlled device

A method comprising:

• receiving, by a device, first input audio data representing first input audio detected by a microphone, the device including a first component associated with a first speech processing system, a second component associated with a second speech processing system, and a third component configured to manage operations of the first component and the second component;
• detecting, in the first input audio data, a representation of a first wakeword corresponding to the first speech processing system;
• in response to detecting the representation of the first wakeword, sending, to the third component, first data comprising a first request for authorization to use the first component to initiate a first dialog session with the first speech processing system;
• receiving, by the first component and from the third component, a first response indicating that the first request has been granted;
• based on the first response, causing the first speech processing system to perform speech processing on at least a portion of the first input audio data;
• receiving speech processing response data from the first speech processing system;
• outputting, by a speaker of the device, first output audio based on the speech processing response data;
• receiving, by the device, second input audio data representing second input audio detected by the microphone;
• detecting, in the second input audio data, a first representation of a second wakeword;
• in response to detecting the first representation of the second wakeword, sending, to the third component, second data comprising a second request for authorization to use the second component to initiate a second dialog session with the second speech processing system;
• determining that the first representation of the second wakeword was detected during an interval of time in which the first output audio was being output;
• receiving, by the second component and from the third component, a second response indicating that the second request has been denied, wherein denying the second request is further based at least in part on the first representation of the second wakeword having been detected during the interval;
• and
• based on the second response, refraining from causing the second speech processing system to perform speech processing on at least a portion of the second input audio data.

A device, comprising:

• a microphone;
• a speaker;
• a first component associated with a first speech processing system;
• a second component associated with a second speech processing system;
• a third component configured to manage operations of the first component and the second component;
• at least one processor;
• and
• at least one memory comprising instructions that, when executed by the at least one processor, cause the device to:
• receive first input audio data representing first input audio detected by the microphone;
• detect, in the first input audio data, a representation of a first wakeword corresponding to the first speech processing system;
• in response to detecting the representation of the first wakeword, send, to the third component, first data comprising a first request for authorization to use the first component to initiate a first dialog session with the first speech processing system;
• receive, by the first component and from the third component, a first response indicating that the first request has been granted;
• based on the first response, cause the first speech processing system to perform speech processing on at least a portion of the first input audio data;
• receive speech processing response data from the first speech processing system;
• output, by the speaker, first output audio based on the speech processing response data;
• receive second input audio data representing second input audio detected by the microphone;
• detect, in the second input audio data, a first representation of a second wakeword;
• in response to detecting the first representation of the second wakeword, send, to the third component, second data comprising a second request for authorization to use the second component to initiate a second dialog session with the second speech processing system;
• determine that the first representation of the second wakeword was detected during an interval of time in which the first output audio was being output;
• receive, by the second component and from the third component, a second response indicating that the second request has been denied, wherein denying the second request is further based at least in part on the first representation of the second wakeword having been detected during the interval;
• and
• based on the second response, refrain from causing the second speech processing system to perform speech processing on at least a portion of the second input audio data.