Speech recognition apparatus, vehicle including the same, and method of controlling the same

• US 9,799,334 B2
• Filed: 10/08/2015
• Issued: 10/24/2017
• Est. Priority Date: 12/17/2014
• Status: Active Grant

• Current Assignee: Hyundai Motor Company (Hyundai Motor Group)
• Original Assignee: Hyundai Motor Company (Hyundai Motor Group)
• Inventors: Lee, Jin Ho
• Primary Examiner(s): Saint Cyr, Leonard
• Application Number:
  US14/879,026
• Publication Number:
  US 20160180846A1
• Time in Patent Office: 747 Days
• Field of Search: 704246, 704247, 704251, 704252
• CPC Class Codes: B60R 16/0373 Voice control in general G10L
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/221 Announcement of recognition results

Abstract

A speech recognition apparatus includes a voice collection unit configured to collect voices, and a controller configured to control recognizing a speech based on signals of the voices collected during a predetermined time period from a voice collection start time, identifying whether a signal is received during the predetermined time period when an operation command corresponding to the recognized speech is not identified, determining that an early utterance occurs upon determination that the signal is received during the predetermined time period, counting a number of times of the speech recognition failure occurring by the early utterance, re-performing speech recognition when the counted number of times is less than a reference number of times, and outputting early utterance habit guidance information when the counted number of times is the same as the reference number of times.

Claims

1. A speech recognition apparatus comprising:

a voice collection unit configured to collect voices; and

a controller configured to control recognizing a speech based on signals of the voices collected by the voice collection unit during a predetermined time period from a voice collection start time, identify whether a signal is received during the predetermined time period when an operation command corresponding to the recognized speech is not identified, generate a No Leading Silence (NLS) flag upon identifying that the signal is received, determine that a speech recognition failure occurs by an early utterance when a number of times NLS flag is generated is the same as a reference number of times, and output information corresponding to the speech recognition failure by the early utterance,

wherein the controller controls re-requesting speech recognition when the generated NLS flag number of times is less than the reference number of times, and outputting early utterance habit guidance information when the generated NLS flag number of times is the same as the reference number of times.

2. The speech recognition apparatus according to claim 1, further comprising an input unit configured to receive a start command for starting speech recognition,

wherein the controller outputs utterance instruction guidance information and an operation command for operating the voice collection unit upon receiving the start command signal for speech recognition.

3. The speech recognition apparatus according to claim 1, wherein when the operation command corresponding to the recognized speech is not identified and a signal is not received during the predetermined time period, the controller outputs information about a command error.

4. The speech recognition apparatus according to claim 1, wherein the controller controls outputting the identified operation command upon identifying the operation command corresponding to the recognized speech.

5. The speech recognition apparatus according to claim 4, wherein when the operation command corresponding to the recognized speech is identified and a signal is received during the predetermined time period, the controller processes the signal received during the predetermined time period as a noise.

6. The speech recognition apparatus according to claim 1, wherein the controller controls outputting a beep sound in the re-requesting of speech recognition.

7. A vehicle comprising:

a speech recognition apparatus; and

at least one device, operation of which is controlled based on an operation command corresponding to a speech recognized by the speech recognition apparatus,

wherein the speech recognition apparatus comprises;

a voice collection unit configured to collect voices; and

a controller configured to control recognizing a speech based on signals of the voices collected by the voice collection unit during a predetermined time period from a voice collection start time, identify whether a signal is received during the predetermined time period when an operation command corresponding to the recognized speech is not identified, generate a No Leading Silence (NLS) flag upon identifying that the signal is received, determine that a speech recognition failure occurs by an early utterance when a number of times NLS flag is generated is the same as a reference number of times, and output information corresponding to the speech recognition failure by the early utterance,

wherein the controller controls re-requesting speech recognition when the generated NLS flag number of times is less than the reference number of times, and output early utterance habit guidance information when the generated NLS flag number of times is the same as the reference number of times.

8. The vehicle according to claim 7, wherein the speech recognition apparatus further comprises an input unit configured to receive a start command for starting speech recognition,

wherein the controller sequentially controls outputting voice guidance information for guiding an utterance instruction process, outputting a beep sound to guide a time for preparing an utterance, and outputting an operation command for operating the voice collection unit upon receiving the start command signal for speech recognition.

9. The vehicle according to claim 7, wherein the controller recognizes a time after a predetermined time period from a time at which the operation command for operating the voice collection unit is output as a voice collection start time, and starts receiving signals of the collected voice.

10. The vehicle according to claim 7, wherein when the operation command corresponding to the recognized speech is identified, the controller controls outputting the identified operation command to the at least one device.

11. The vehicle according to claim 7, wherein when the operation command corresponding to the recognized speech is identified and a signal is received during the predetermined time period, the controller processes the signal received during the predetermined time period as a noise, and

when the operation command corresponding to the recognized speech is not identified and a signal is not received during the predetermined time period, the controller determines that a command error occurs and controls outputting information about the command error to the at least one device.

12. The vehicle according to claim 7, wherein the controller controls outputting the early utterance habit guidance information by estimating occurrence of no leading silence (NLS) by early utterance when the generated NLS flag number of times is one and confirming occurrence of NLS by early utterance when the generated NLS flag number of times is two.

13. The vehicle according to claim 7, further comprising a display unit configured to display the early utterance habit guidance information.

14. A method of controlling a speech recognition apparatus configured to receive a command to control operation of a plurality of devices installed in a vehicle, the method comprising:

determining whether a start command signal for speech recognition is received;

operating a voice collection unit upon determination that the start command signal for speech recognition is received;

determining whether a predetermined time period has lapsed from a voice collection start time;

recognizing a speech based on signals of voices collected from a time after the predetermined time period;

identifying an operation command corresponding to the recognized speech;

identifying whether a signal is received during the predetermined time period when the operation command corresponding to the recognized speech is not identified;

generating a No Leading Silence (NLS) flag upon identifying that the signal is received, determining that a speech recognition failure occurs by an early utterance when a number of times NLS flag is generated is the same as a reference number of times;

outputting information corresponding to the speech recognition failure by the early utterance; and

counting a number of times of the speech recognition failure by the early utterance, outputting a beep sound to re-perform speech recognition when the generated NLS flag number of times is less than the reference number of times, and outputting early utterance habit guidance information when the generated NLS flag number of times is the same as the reference number of times.

15. The method according to claim 14, further comprising outputting voice guidance information for guiding an utterance instruction process upon receiving the start command signal for speech recognition;

outputting a beep sound to guide a time for preparing an utterance when the outputting of the voice guidance information is complete;

outputting an operation command for operating the voice collection unit when the outputting of the beep sound is complete; and

recognizing a time after a predetermined time period from a time at which the operation command for operating the voice collection unit is output as a voice collection start time.

16. The method according to claim 14, further comprising outputting the identified operation command to one of the plurality of devices installed in the vehicle when the operation command corresponding to the recognized speech is identified.

17. The method according to claim 14, further comprising processing a signal received during the predetermined time period as a noise when the operation command corresponding to the recognized speech is identified and the signal is received during the predetermined time period.