Systems and methods for selective wake word detection using neural network models
Systems and methods for media playback via a media playback system include capturing sound data via a network microphone device and identifying a candidate wake word in the sound data. Based on identification of the candidate wake word in the sound data, the system selects a first wake-word engine from a plurality of wake-word engines. Via the first wake-word engine, the system analyzes the sound data to detect a confirmed wake word, and, in response to detecting the confirmed wake word, transmits a voice utterance of the sound data to one or more remote computing devices associated with a voice assistant service.
1. A method comprising:
capturing sound data via a network microphone device;
identifying, using a keyword spotting algorithm, a candidate wake word in the sound data;
based on identification of the candidate wake word in the sound data via the keyword spotting algorithm, selecting a first wake-word detection algorithm from among a plurality of wake-word detection algorithms stored on the network microphone device, wherein the first wake-word detection algorithm is associated with a first voice assistant service and another of the plurality of wake-word detection algorithms is associated with a second voice assistant service different from the first;
after selecting the first wake-word detection algorithm, and without using another of the plurality of wake-word detection algorithms, using the first wake-word detection algorithm to analyze the sound data to confirm detection of the candidate wake word identified via the keyword spotting algorithm, wherein the first wake-word detection algorithm is configured to determine whether the candidate wake word is present in the sound data with a higher accuracy than the keyword spotting algorithm; and
in response to confirming the detection of the candidate wake word, transmitting a voice utterance of the sound data to one or more remote computing devices associated with the first voice assistant service.
2. The method of claim 1 , wherein analyzing the sound data to confirm detection of the candidate wake word using the using the first wake-word detection algorithm, and without using another of the wake-word detection algorithms comprises:
activating a first wake-word engine to process the sound data using the first wake-word detection algorithm while a second wake-word engine configured to process the sound data using another of the wake-word detection algorithms is in an inactive state.
3. The method of claim 1 , wherein identifying the candidate wake word comprises determining a probability that the candidate wake word is present in the sound data.
4. The method of claim 1 , wherein the first wake-word detection algorithm is associated with the candidate wake word, and wherein another of the plurality of wake-word detection algorithms is associated with one or more additional wake words.
5. The method of claim 1 , wherein identifying the candidate wake word comprises applying a neural network model to the sound data.
6. The method of claim 5 , wherein the neural network model comprises a compressed neural network model stored locally on the network microphone device.
7. The method of claim 1 , further comprising, after transmitting the voice utterance, receiving, via the network microphone device, a selection of media content related to the voice utterance.
8. The method of claim 1 , wherein the plurality of wake-word detection algorithms comprises:
the first wake-word detection algorithm; and
a second wake-word detection algorithm configured to perform a local function of the network microphone device.
9. A network microphone device, comprising:
one or more processors;
at least one microphone; and
tangible, non-transitory, computer-readable media storing instructions executable by one or more processors to cause the network microphone device to perform operations comprising:
capturing sound data via the one or more microphones;
identifying, using a keyword spotting algorithm, a candidate wake word in the sound data;
based on identification of the candidate wake word in the sound data via the keyword spotting algorithm, selecting a first wake-word detection algorithm from among a plurality of wake-word detection algorithms stored on the network microphone device, wherein the first wake-word detection algorithm is associated with a first voice assistant service and another of the plurality of wake-word detection algorithms is associated with a second voice assistant service different from the first;
after selecting the first wake-word detection algorithm, and without using another of the plurality of wake-word detection algorithms, using the first wake-word detection algorithm to analyze the sound data to confirm detection of the candidate wake word identified via the keyword spotting algorithm, wherein the first wake-word detection algorithm is configured to determine whether the candidate wake word is present in the sound data with a higher accuracy than the keyword spotting algorithm; and
in response to confirming the detection of the candidate wake word, transmitting a voice utterance of the sound data to one or more remote computing devices associated with the first voice assistant service.
10. The network microphone device of claim 9 , wherein analyzing the sound data to confirm detection of the candidate wake word using the using the first wake-word detection algorithm, and without using another of the wake-word detection algorithms comprises:
activating a first wake-word engine to process the sound data using the first wake-word detection algorithm while a second wake-word engine configured to process the sound data using another of the wake-word detection algorithms is in an inactive state.
11. The network microphone device of claim 9 , wherein identifying the candidate wake word comprises determining a probability that the candidate wake word is present in the sound data.
12. The network microphone device of claim 9 , wherein the first wake-word detection algorithm is associated with the candidate wake word, and wherein another of the plurality of wake-word detection algorithms is associated with one or more additional wake words.
13. The network microphone device of claim 9 , wherein identifying the candidate wake word comprises applying a neural network model to the sound data.
14. The network microphone device of claim 9 , further comprising, after transmitting the voice utterance, receiving, via the network microphone device, a selection of media content related to the voice utterance.
15. The network microphone device of claim 9 , wherein the plurality of wake-word detection algorithms comprises:
the first wake-word detection algorithm; and
a second wake-word detection algorithm configured to perform a local function of the network microphone device.
16. Tangible, non-transitory, computer-readable media storing instructions executable by one or more processors to cause a network microphone device to perform operations comprising:
capturing sound data via the network microphone device;
identifying, using a keyword spotting algorithm, a candidate wake word in the sound data;
based on identification of the candidate wake word in the sound data via the keyword spotting algorithm, selecting a first wake-word detection algorithm from among a plurality of wake-word detection algorithms stored on the network microphone device, wherein the first wake-word detection algorithm is associated with a first voice assistant service and another of the plurality of wake-word detection algorithms is associated with a second voice assistant service different from the first;
after selecting the first wake-word detection algorithm, and without using another of the plurality of wake-word detection algorithms, using the first wake-word detection algorithm to analyze the sound data to confirm detection of the candidate wake word identified via the keyword spotting algorithm, wherein the first wake-word detection algorithm is configured to determine whether the candidate wake word is present in the sound data with a higher accuracy than the keyword spotting algorithm; and
in response to confirming the detection of the candidate wake word, transmitting a voice utterance of the sound data to one or more remote computing devices associated with the first voice assistant service.
17. The computer-readable media of claim 16 , wherein analyzing the sound data to confirm detection of the candidate wake word using the using the first wake-word detection algorithm, and without using another of the wake-word detection algorithms comprises:
activating a first wake-word engine to process the sound data using the first wake-word detection algorithm while a second wake-word engine configured to process the sound data using another of the wake-word detection algorithms is in an inactive state.
18. The computer-readable media of claim 16 , wherein the first wake-word detection algorithm is associated with the candidate wake word, and wherein another of the plurality of wake-word detection algorithms is associated with one or more additional wake words.
19. The computer-readable media of claim 16 , wherein identifying the candidate wake word comprises applying a neural network model to the sound data.