Voice detection optimization based on selected voice assistant service
Systems and methods for optimizing voice detection via a network microphone device (NMD) based on a selected voice-assistant service (VAS) are disclosed herein. In one example, the NMD detects sound via individual microphones and selects a first VAS to communicate with the NMD. The NMD produces a first sound-data stream based on the detected sound using a spatial processor in a first configuration. Once the NMD determines that a second VAS is to be selected over the first VAS, the spatial processor assumes a second configuration for producing a second sound-data stream based on the detected sound. The second sound-data stream is then transmitted to one or more remote computing devices associated with the second VAS.
1. A method, comprising:
detecting sound via individual microphones of a network microphone device (NMD);
selecting a first voice assistant service (VAS) to communicate with the NMD;
producing a first sound-data stream based on the detected sound using a spatial processor, wherein the spatial processor is in a first configuration when the first VAS is selected;
determining that a second VAS is to be selected over the first VAS based on the detected sound;
selecting the second VAS to communicate with the NMD and selection of the first VAS;
producing a second sound-data stream based on the detected sound using the spatial processor, wherein the spatial processor is in a second configuration when the second VAS is selected; and
transmitting the second sound-data stream to a remote computing device associated with the second VAS.
2. The method of claim 1 , wherein the spatial processor is configured to process first and second numbers of channels of the detected sound in the first and second configurations, respectively.
3. The method of claim 1 , wherein:
the spatial processor comprises an algorithm having filter coefficients, and
the filter coefficients are different in the first configuration than in the second configuration.
4. The method of claim 1 , wherein the spatial processor utilizes a beamforming algorithm and a multi-channel wiener filter algorithm in the first and second configurations, respectively.
5. The method of claim 1 , further comprising refraining from transmitting the first or second sound-data stream until the determination is made.
6. The method of claim 1 , wherein determining that the second VAS is to be selected over the first VAS comprises identifying a wake-word associated with the second VAS in the first sound-data stream.
7. The method of claim 1 , further comprising:
causing, by a pipeline selector, the spatial processor to use a particular configuration corresponding to the selected VAS.
8. The method of claim 7 , wherein the pipeline selector is further configured to cause the spatial processor to modify at least one of:
a fixed gain of one or more microphones;
a wake-word sensitivity parameter; or
noise-reduction parameters.
9. The method of any claim 1 , wherein the first configuration of the spatial processor is selected by a manual user configuration.
10. The method of claim 1 , wherein the first configuration of the spatial processor is associated with the first VAS.
11. The method of claim 1 , further comprising:
transmitting the second sound stream to the second VAS;
receiving, from the second VAS, a command to be executed by the NMD.
12. The method of claim 1 , further comprising;
transmitting the second sound stream to the second VAS;
receiving, from the second VAS, a response to be output by the NMD.
13. The method of claim 1 , further comprising:
capturing metadata associated with the sound data in a lookback buffer;
providing the metadata to at least one network device to determine at least one characteristic of the detected sound based on the metadata;
after providing the metadata, receiving, from the at least one network device, a response including an instruction, based on the determined characteristic, to modify at least one performance parameter of the network microphone device; and
modifying the at least one performance parameter based on the instruction.
14. A non-transitory computer-readable medium comprising instructions for adjusting performance of a network microphone device (NMD), the instructions, when executed by a processor, causing the processor to perform the method of claim 1 .
15. A network microphone device (NMD) comprising:
one or more processors;
a microphone array comprising a plurality of individual microphones; and
a computer-readable medium according to claim 14 .