Method for recognizing a voice context for a voice control function, method for ascertaining a voice control signal for a voice control function, and apparatus for executing the method

• US 9,865,258 B2
• Filed: 05/17/2016
• Issued: 01/09/2018
• Est. Priority Date: 06/08/2015
• Status: Active Grant

• Current Assignee: Robert Bosch GmbH
• Original Assignee: Robert Bosch GmbH
• Inventors: Dreuw, Philippe
• Primary Examiner(s): PULLIAS, JESSE SCOTT
• Application Number:
  US15/156,721
• Publication Number:
  US 20160358604A1
• Time in Patent Office: 602 Days
• Field of Search: 704231-257, 704270-275
• CPC Class Codes: G06F 3/012 Head tracking input arrangements
• CPC Class Codes: G06F 3/013 Eye tracking input arrangements G06F3/015 takes precedence
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/227 of the speaker; Human-factor methodology

Abstract

A method for recognizing a voice context for a voice control function in a vehicle. The method encompasses reading in a gaze direction datum regarding a current gaze direction of an occupant of the vehicle; allocating the gaze direction datum to a viewing zone in an interior of the vehicle in order to obtain a viewing zone datum regarding a viewing zone currently being viewed by the occupant; and determining, by utilization of the viewing zone datum, a voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed.

Claims

1. A method for ascertaining a voice control signal for a voice control function in a vehicle, the method comprising:

reading in a voice context datum, the voice context datum being determined by;

reading in a gaze direction datum regarding a current gaze direction of an occupant of the vehicle,

allocating the gaze direction datum to a viewing zone of a plurality of viewing zones, each of the plurality of viewing zones corresponding to a different region, in an interior of the vehicle to obtain a viewing zone datum regarding a viewing zone currently being viewed by the occupant, and

determining, by using the viewing zone datum, the voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed,

reading in a further gaze direction datum regarding a further current gaze direction of a further occupant of the vehicle;

allocating the further gaze direction datum to a further viewing zone in the interior of the vehicle to obtain a further viewing zone datum regarding a further viewing zone currently being viewed by the further occupant; and

determining, by using the further viewing zone datum, a further voice context datum regarding a further predetermined voice context allocated to the further viewing zone currently being viewed;

reading in an audio signal; and

ascertaining the voice control signal by using the voice context datum and the audio signal;

wherein the allocating is executed when a predetermined viewing time threshold value is exceeded in the reading-in, the viewing time representing a time during which the viewing zone is being viewed by the occupant.

2. The method as recited in claim 1, wherein in the allocating step, the viewing zone is selected, by using the gaze direction datum, from a plurality of viewing zones.

3. The method as recited in claim 1, further comprising:

selecting, by using the voice context datum, a predetermined voice model for the voice control function.

4. The method as recited in claim 3, further comprising:

allocating the occupant, by using the gaze direction datum, to an occupant category from among a plurality of occupant categories, the voice context datum being further determined in the determining step by using the occupant category allocated to the occupant.

5. The method as recited in claim 4, wherein in the allocating the occupant to an occupant category step, the occupant is identified as an individual person.

6. The method as recited in claim 1, further comprising:

reading in a head posture datum regarding a current head posture of the occupant of the vehicle; and

allocating the head posture datum to the viewing zone in the interior of the vehicle to obtain an additional viewing zone datum regarding a viewing zone currently being viewed by the occupant, the voice context datum being further determined, in the step of determining a voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed, by using the additional viewing zone datum.

7. The method as recited in claim 1, further comprising:

reading in a lip position datum regarding a current lip position of the occupant of the vehicle; and

allocating the lip position datum to the viewing zone in the interior of the vehicle to obtain a second additional viewing zone datum regarding a viewing zone currently being viewed by the occupant, the voice context datum being further determined, in the step of determining a voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed, by using the second additional viewing zone datum.

8. The method as recited in claim 1, wherein in reading in of the audio signal step, the audio signal is read in at least partly from a buffer, the buffer being embodied to store at least the audio signal for a predetermined time period.

9. A method for recognizing a voice context for a voice control function in a vehicle, the method comprising:

reading in a gaze direction datum regarding a current gaze direction of an occupant of the vehicle;

allocating the gaze direction datum to a viewing zone in an interior of the vehicle to obtain a viewing zone datum regarding a viewing zone currently being viewed by the occupant;

determining, by using the viewing zone datum, a voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed;

reading in a further gaze direction datum regarding a further current gaze direction of a further occupant of the vehicle;

allocating the further gaze direction datum to a further viewing zone in the interior of the vehicle to obtain a further viewing zone datum regarding a further viewing zone currently being viewed by the further occupant; and

determining, by using the further viewing zone datum, a further voice context datum regarding a further predetermined voice context allocated to the further viewing zone currently being viewed.

10. An apparatus for recognizing a voice context for a voice control function in a vehicle, comprising:

a read-in arrangement to read in a gaze direction datum regarding a current gaze direction of an occupant of the vehicle;

an allocating arrangement to allocate the gaze direction datum to a viewing zone of a plurality of viewing zones, each of the plurality of viewing zones corresponding to a different region, in an interior of the vehicle to obtain a viewing zone datum regarding a viewing zone currently being viewed by the occupant; and

a determining arrangement to determine, by using the viewing zone datum, a voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed;

wherein the allocating is executed when a predetermined viewing time threshold value is exceeded in the reading-in, the viewing time representing a time during which the viewing zone is being viewed by the occupant,

wherein the reading in arrangement is further configured to read in a further gaze direction datum regarding a further current gaze direction of a further occupant of the vehicle, wherein the allocating arrangement is further configured to allocate the further gaze direction datum to a further viewing zone in the interior of the vehicle to obtain a further viewing zone datum regarding a further viewing zone currently being viewed by the further occupant; and

wherein the determining arrangement is further configured to determine, by using the further viewing zone datum, a further voice context datum regarding a further predetermined voice context allocated to the further viewing zone currently being viewed.

11. An apparatus for ascertaining a voice control signal for a voice control function in a vehicle, comprising:

a first reading arrangement to read in a voice context datum, the voice context datum being determined by performing the following;

reading in, via a reading in arrangement, a gaze direction datum regarding a current gaze direction of an occupant of the vehicle,

allocating, via an allocating arrangement, the gaze direction datum to a viewing zone of a plurality of viewing zones, each of the plurality of viewing zones corresponding to a different region, in an interior of the vehicle to obtain a viewing zone datum regarding a viewing zone currently being viewed by the occupant, and

determining, via a determining arrangement, by using the viewing zone datum, the voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed,

a second reading arrangement to read in an audio signal; and

an ascertaining arrangement to ascertain the voice control signal by using the voice context datum and the audio signal;

wherein the allocating is executed when a predetermined viewing time threshold value is exceeded in the reading-in, the viewing time representing a time during which the viewing zone is being viewed by the occupant,

wherein the reading in arrangement is further configured to read in a further gaze direction datum regarding a further current gaze direction of a further occupant of the vehicle, wherein the allocating arrangement is further configured to allocate the further gaze direction datum to a further viewing zone in the interior of the vehicle to obtain a further viewing zone datum regarding a further viewing zone currently being viewed by the further occupant; and

wherein the determining arrangement is further configured to determine, by using the further viewing zone datum, a further voice context datum regarding a further predetermined voice context allocated to the further viewing zone currently being viewed.

12. A machine-readable non-transitory storage medium storing a computer program, which is executable by a processor, comprising:

a program code arrangement having program code for recognizing a voice context for a voice control function in a vehicle, by performing the following;

reading in, via a reading in arrangement, a gaze direction datum regarding a current gaze direction of an occupant of the vehicle;

allocating, via an allocating arrangement, the gaze direction datum to a viewing zone of a plurality of viewing zones, each of the plurality of viewing zones corresponding to a different region, in an interior of the vehicle to obtain a viewing zone datum regarding a viewing zone currently being viewed by the occupant;

determining, via a determining arrangement, by using the viewing zone datum, a voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed;

reading in, via the reading in arrangement, a further gaze direction datum regarding a further current gaze direction of a further occupant of the vehicle;

allocating, via the allocating arrangement, the further gaze direction datum to a further viewing zone in the interior of the vehicle to obtain a further viewing zone datum regarding a further viewing zone currently being viewed by the further occupant; and

determining, via the determining arrangement, by using the further viewing zone datum, a further voice context datum regarding a further predetermined voice context allocated to the further viewing zone currently being viewed;

wherein the allocating is executed when a predetermined viewing time threshold value is exceeded in the reading-in, the viewing time representing a time during which the viewing zone is being viewed by the occupant.

13. A machine-readable non-transitory storage medium storing a computer program, which is executable by a processor, comprising:

a program code arrangement having program code for ascertaining a voice control signal for a voice control function in a vehicle, by performing the following;

a first reading arrangement to read in a voice context datum, the voice context datum being determined by performing the following;

reading in, via a reading in arrangement, a gaze direction datum regarding a current gaze direction of an occupant of the vehicle,

allocating, via an allocating arrangement, the gaze direction datum to a viewing zone of a plurality of viewing zones, each of the plurality of viewing zones corresponding to a different region, in an interior of the vehicle to obtain a viewing zone datum regarding a viewing zone currently being viewed by the occupant, and

determining, via a determining arrangement, by using the viewing zone datum, the voice context datum regarding a predetermined voice context allocated to the viewing zone currently being viewed,

a second reading arrangement to read in an audio signal; and

an ascertaining arrangement to ascertain the voice control signal by using the voice context datum and the audio signal;

wherein the allocating is executed when a predetermined viewing time threshold value is exceeded in the reading-in, the viewing time representing a time during which the viewing zone is being viewed by the occupant,

wherein the reading in arrangement is further configured to read in a further gaze direction datum regarding a further current gaze direction of a further occupant of the vehicle, wherein the allocating arrangement is further configured to allocate the further gaze direction datum to a further viewing zone in the interior of the vehicle to obtain a further viewing zone datum regarding a further viewing zone currently being viewed by the further occupant; and

wherein the determining arrangement is further configured to determine, by using the further viewing zone datum, a further voice context datum regarding a further predetermined voice context allocated to the further viewing zone currently being viewed.