Co-talker nulling for automatic speech recognition systems

• US 9,431,013 B2
• Filed: 11/07/2013
• Issued: 08/30/2016
• Est. Priority Date: 11/07/2013
• Status: Active Grant

• Current Assignee: Continental Automotive Systems, Inc. (Continental AG)
• Original Assignee: Continental Automotive Systems, Inc. (Continental AG)
• Inventors: Reuter, Mike
• Primary Examiner(s): Desir, Pierre-Louis
• Assistant Examiner(s): Sharma, Neeraj
• Application Number:
  US14/074,609
• Publication Number:
  US 20150127338A1
• Time in Patent Office: 1,027 Days
• Field of Search: 704/233, 704/226, 704/251, 704/275, 381/92, 381/66, 348/148
• US Class Current: 1/1
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 2021/02165 Two microphones, one receiving mainly the noise signal and the other one mainly the speech signal
• CPC Class Codes: G10L 21/0316 by changing the amplitude
• CPC Class Codes: H04R 1/08 Mouthpieces; Microphones; Attachments therefor
• CPC Class Codes: H04R 1/406 microphones
• CPC Class Codes: H04R 2430/20 Processing of the output signals of the acoustic transducers of an array for obtaining a desired directivity characteristic H04R2203/12 takes precedence
• CPC Class Codes: H04R 2499/13 Acoustic transducers and sound field adaptation in vehicles
• CPC Class Codes: H04R 3/005 for combining the signals of two or more microphones specially adapted for hearing aids H04R25/407

Abstract

Speech from a driver and speech from a passenger in a vehicle is selected directionally using two or more microphones. Samples of speech from a driver and picked up by a first microphone are delayed until samples of the speech picked up by a second microphone is in phase with the speech picked up by the first microphone. Samples of a passenger'"'"'s speech and picked up by the second microphone are delayed until samples of the passenger'"'"'s speech picked up by the first microphone are out-of-phase with the speech picked up by the second microphone.

Claims

1. In a vehicle having a driver and a passenger and first and second microphones, both microphones being capable of detecting sounds from both the passenger and driver, a method for selectively detecting sounds from the driver and suppressing sounds from the passenger, the method comprising:

selectively detecting at the first microphone, a first sample of a sound from the driver, the first sample of the sound from the driver comprising a first frequency component, the first sample of the sound from the driver being a sample of the first frequency component at a first phase angle;

selectively detecting at the second microphone, which is spatially separated from the first microphone by a distance of at least twelve inches, a second sample of the sound from the driver, the second sample of the sound from the driver also comprising the first frequency component, said second sample being a sample of the first frequency component having a second and different phase angle;

storing the first sample of the sound from the driver for a first predetermined length of time to provide a delayed first sample of the sound from the driver, the first predetermined length of time selected such that the phase angles of the delayed first sample of the sound from the driver and of the second sample of the sound from the driver are made to be substantially equal to each other;

combining the delayed first sample and the second sample of the sound from the driver to provide a first constructive interference;

selectively detecting at the first microphone, a first sample of a sound from the passenger, the first sample of the sound from the passenger comprising a second frequency component, the first sample of the sound from the passenger being a sample of the second frequency component at a third phase angle;

selectively detecting at the second microphone, a second sample of the sound from the passenger, the second sample of the sound from the passenger comprising the second frequency component, the second sample being a sample of the second frequency component at a fourth phase angle;

storing the first sample of the sound from the passenger for a second predetermined length of time to provide a delayed first sample of the sound from the passenger, the second predetermined length of time being selected such that the phase angle of the delayed first sample of the sound from the passenger and the fourth phase angle are substantially equal to each other;

combining the delayed first sample and the second sample of the sound from the passenger to provide a second constructive interference, the second predetermined length of time being additionally selected to determine a direction from which sound from the passenger will be directionally detected by said first and second microphones;

determining the location of the driver in the vehicle as corresponding to where a first directionally selective microphone beam is directed and whereat the amplitude of the constructive interference of the first and second sound samples from the driver is greatest;

determining the location of the passenger in the vehicle as corresponding to the direction of a second directionally selective microphone beam directed at the passenger, whereat the amplitude of the constructive interference of the first and second sound samples from the passenger is greatest; and

filtering sounds to be provided to a speech recognizer by selectively detecting the sounds of the driver and suppressing the sounds from the passenger, by subtracting the second constructive interference as per the determined location of the passenger, from the first constructive interference as per the determined location of the driver.

2. The method of claim 1, wherein the second predetermined length of time is substantially equal to an integer multiple of the length of time required for one wavelength of the first frequency component to travel between the first microphone and second microphone.

3. The method of claim 1, wherein selectively detecting samples of sounds at the first and second microphones directionally selects sounds from at least one of the driver and passenger.

4. An apparatus for a vehicle having a driver and a passenger and first and second microphones, both microphones being capable of detecting sounds from both the passenger and driver, the apparatus configured to selectively detect sounds from the driver and suppress sounds from the passenger, the apparatus comprising:

a digital signal processor coupled to the first and second microphones;

a non-transitory memory device coupled to the digital signal processor and storing program instructions, which when executed cause the digital signal processor to;

selectively detecting at a first microphone, a first sample of a sound from the driver, the first sample of the sound from the driver being a sample of a first frequency component at a first phase angle;

selectively detecting at a second microphone, which is spatially separated from the first microphone by a distance of at least twelve inches, a second sample of the sound from the driver, said second sample being a sample of the first frequency component at a second phase angle;

storing the first sample for a first predetermined length of time to provide a delayed first sample of sound from the driver, the first predetermined length of time selected such that the phase angles of the delayed first sample of sound from the driver and the second phase angle are substantially equal to each other; and

combining the delayed first sample and the second sample of sound from the driver to provide a first constructive interference;

selectively detecting at the first microphone, a first sample of a sound from the passenger, the first sample of the sound from the passenger being a sample of a third frequency component at a third phase angle;

selectively detecting at the second microphone, a second sample of the sound from the passenger, said second sample being a sample of the third frequency component at a fourth phase angle;

storing the first sample of sound from the passenger for a predetermined second length of time to provide a delayed first sample of sound from the passenger, the predetermined second length of time being selected such that the phase angle of the delayed first sample of sound from the passenger and the fourth phase angle are substantially equal to each other; and

combining the delayed first sample and the second sample of the sound from the passenger to provide a second constructive interference, the second predetermined length of time being additionally selected to determine a direction from which sound from the passenger will be directionally detected by said first and second microphones;

determining the location of the driver in the vehicle as corresponding to where a first directionally selective microphone beam is directed and whereat the amplitude of the constructive interference of the first and second sound samples from the driver is greatest;

determining the location of the passenger in the vehicle as corresponding to the direction of a second directionally selective microphone beam directed at the passenger, whereat the amplitude of the constructive interference of the first and second sound samples from the passenger is greatest; and

filtering sounds to be provided to a speech recognizer by selectively detecting the sounds of the driver and suppressing the sounds from the passenger, by subtracting the second constructive interference as per the determined location of the passenger, from the first constructive interference as per the determined location of the driver.

5. The apparatus of claim 4, wherein the digital signal processor is configured such that the first and second predetermined lengths of time are substantially equal to an integer multiple of the length of time required for one wavelength of the first frequency component to travel between the first microphone and second microphone.

6. The apparatus of claim 4, wherein the digital signal processor is configured such that selectively detecting samples of sounds at the first and second microphones directionally selects sounds from at least one of the driver and passenger.