Dynamic microphone system for autonomous vehicles
Devices, systems and processes for a dynamic microphone system that enhances the passenger experience in autonomous vehicles are described. One example method for enhancing a passenger experiences includes generating, using an artificial intelligence algorithm, a plurality of filters based on a plurality of stored waveforms previously recorded by each of one or more passengers and a plurality of recordings of one or more noise sources, capturing voice commands from at least one of the one or more passengers inside the autonomous vehicle, generating voice commands with reduced distortion based on processing the voice commands using the plurality of filters, and instructing, based on the voice commands with reduced distortion, the autonomous vehicle to perform one or more actions.
1. A system for use in a vehicle, comprising:
an audio processor; and
a plurality of microphones coupled to the audio processor, wherein the plurality of microphones is configured to capture voice commands from one or more passengers inside of the vehicle,
wherein the audio processor is configured to:
perform speaker identification to determine which voice commands correspond to which of the one or more passengers,
process the voice commands to generate reduced distortion voice commands, and
perform speech recognition on the reduced distortion voice commands to
determine one or more requests made by the one or more passengers,
wherein the audio processor is configured to process the voice commands to generate reduced distortion voice commands using a filter generated by executing an artificial intelligence model using the voice commands and an identification of a current configuration of the vehicle as inputs to the artificial intelligence model,
wherein the artificial intelligence model comprises an artificial neural network trained using training inputs that include synthetic noisy speech,
wherein the synthetic noisy speech comprises noise and voice recordings,
wherein the noise comprises recordings of internal noise sources recorded for different configurations of the vehicle, and
wherein the training inputs for the artificial neural network further include an indication of a configuration of the vehicle when a particular noise recording was captured.
2. The system of claim 1 ,
wherein the audio processor is configured to process the voice commands to generate reduced distortion voice commands by comparing the voice commands to one or more stored waveforms previously recorded for a first passenger of the one or more passengers.
3. The system of claim 2 ,
wherein the audio processor is configured to process the voice commands to generate reduced distortion voice commands by determining expected voice commands based on a result of comparing the voice commands to the one or more stored waveforms.
4. The system of claim 3 ,
wherein the audio processor is configured to process the voice commands to generate reduced distortion voice commands by identifying acoustic anomalies in the voice commands that represent deviations from the expected voice commands, and
wherein the acoustic anomalies are due to environmental factors, a location of the first passenger in the autonomous vehicle, a volume level or intensity of the voice commands, and/or ambient noise due to other factors.
5. The system of claim 4 ,
wherein the audio processor is configured to process the voice commands to generate reduced distortion voice commands by removing the acoustic anomalies from the voice commands to generate the reduced-distortion voice commands.
6. The system of claim 1 ,
wherein the synthetic noisy speech comprises the noise superimposed on the voice recordings.
7. The system of claim 1 ,
wherein the noise comprises recordings of environmental noise sources recorded for different configurations of the vehicle.
8. The system of claim 1 ,
wherein the voice recordings comprise voice recordings of multiple candidate passengers, and
wherein the voice recordings of the multiple candidate passengers contain minimal external acoustic interferences.
9. The system of claim 1 ,
wherein the indication of the configuration of the vehicle when a particular noise recording was captured includes an indication of a type of car in which the particular noise recording was captured.
10. The system of claim 1 ,
wherein the indication of the configuration of the vehicle when a particular noise recording was captured includes an indication of a configuration of a seat in the vehicle when the particular noise recording was captured.
11. A method for use in a vehicle, comprising:
capturing voice commands from one or more passengers inside of the vehicle;
performing speaker identification to determine which voice commands correspond to which of the one or more passengers;
processing the voice commands to generate reduced distortion voice commands using a filter generated by executing an artificial intelligence model using the voice commands and an identification of a current configuration of the vehicle as inputs to the artificial intelligence model; and
performing speech recognition on the reduced distortion voice commands to determine one or more requests made by the one or more passengers,
wherein the artificial intelligence model comprises an artificial neural network trained using training inputs that include synthetic noisy speech,
wherein the synthetic noisy speech comprises noise and voice recordings,
wherein the noise comprises recordings of internal noise sources recorded for different configurations of the vehicle, and
wherein the training inputs further include an indication of a configuration of the vehicle when a particular noise recording was captured.
12. The method of claim 11 ,
wherein processing the voice commands to generate reduced distortion voice commands comprises comparing the voice commands to one or more stored waveforms previously recorded for a first passenger of the one or more passengers.
13. The method of claim 12 ,
wherein processing the voice commands to generate reduced distortion voice commands comprises determining expected voice commands based on a result of comparing the voice commands to the one or more stored waveforms.
14. The method of claim 13 ,
wherein processing the voice commands to generate reduced distortion voice commands comprises identifying acoustic anomalies in the voice commands that represent deviations from the expected voice commands, and
wherein the acoustic anomalies are due to environmental factors, a location of the first passenger in the autonomous vehicle, a volume level or intensity of the voice commands, and/or ambient noise due to other factors.
15. The method of claim 14 ,
wherein processing the voice commands to generate reduced distortion voice commands comprises removing the acoustic anomalies from the voice commands to generate the reduced-distortion voice commands.
16. The method of claim 11 ,
wherein the synthetic noisy speech comprises the noise superimposed on the voice recordings.
17. The method of claim 11 ,
wherein the noise comprises recordings of environmental noise sources recorded for different configurations of the vehicle.
18. The method of claim 11 ,
wherein the voice recordings comprise voice recordings of multiple candidate passengers, and
wherein the voice recordings of the multiple candidate passengers contain minimal external acoustic interferences.
19. The method of claim 11 ,
wherein the indication of the configuration of the vehicle when a particular noise recording was captured includes an indication of a type of car in which the particular noise recording was captured.
20. The method of claim 11 ,
wherein the indication of the configuration of the vehicle when a particular noise recording was captured includes an indication of a configuration of a seat in the vehicle when the particular noise recording was captured.