Baseband modem for speech recognition and mobile communication terminal using the same

• US 20060053011A1
• Filed: 09/07/2005
• Published: 03/09/2006
• Est. Priority Date: 09/07/2004
• Status: Active Grant

• Current Assignee: LG Electronics, Inc. (LG Corporation)
• Original Assignee: LG Electronics, Inc. (LG Corporation)
• Inventors: Kim, Chan Woo
• Granted Patent: US 7,593,853 B2
• Time in Patent Office: Days
• US Class Current: 704/236
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: H04M 2201/40 using speech recognition

Abstract

A baseband modem and method for voice recognition and a mobile communication terminal using the baseband modem and method are disclosed. A speech recognition rate may be increased by selecting a sampling rate suitable for speech recognition and portions of the speech recognition process may be implemented in hardware. The present invention includes an audio codec modulating a received voice signal using either a sampling rate for speech recognition or a sampling rate for voice communication. A feature vector extraction block extracts one or more feature vectors from the modulated voice signal and a speech recognition block performs speech recognition using an extracted feature vector when the voice signal is determined as a voice command. A vocoder vocodes an output of the audio codec when the voice signal is determined as voice communication.

Claims

1. A baseband modem, comprising:

an audio codec adapted to modulate a voice signal using one of a first sampling rate and a second sampling rate;

means for speech recognition; and

means for speech encoding,

wherein the audio codec encodes the voice signal using the first sampling rate and speech recognition means performs speech recognition of the encoded voice signal if the voice signal is a voice command and the audio codec encodes the voice signal using the second sampling rate and the speech encoding means performs vocoding of the encoded voice signal if the voice signal is voice communication.

2. The baseband modem of claim 1, wherein the speech recognition means further comprises:

a feature vector extraction block adapted to extract at least one feature vector from the encoded voice signal; and

a speech recognition block adapted to perform speech recognition using the at least one feature vector extracted by the feature vector extraction block.

3. The baseband modem of claim 2, wherein the speech recognition means further comprises a buffer adapted to store the encoded voice signal.

4. The baseband modem of claim 3, wherein the feature vector extraction block is further adapted to extract the at least one feature vector from data stored in the buffer.

5. The baseband modem of claim 3, wherein the buffer comprises a ping-pong buffer.

6. The baseband modem of claim 2, wherein the speech recognition means further comprises a buffer adapted to store the at least one feature vector extracted from the encoded voice signal.

7. The baseband modem of claim 2, wherein the feature vector extraction block is implemented in hardware.

8. The baseband modem of claim 2, wherein the feature vector extraction block is implemented in software.

9. The baseband modem of claim 2, further comprising a controller adapted to determine that the voice signal is one of a voice command and voice communication and to power on registers of the feature vector extraction block and speech recognition block if the voice signal is a voice command and power off registers of the feature vector extraction block and speech recognition block if the voice signal is voice communication.

10. The base modem of claim 1, wherein the speech encoding means further comprises a vocoder adapted to vocode the encloded voice signal.

11. The baseband modem of claim 1, further comprising a controller adapted to determine that the voice signal is one of a voice command and voice communication.

12. The baseband modem of claim 11, wherein the controller is further adapted to determine the sampling rate used by the audio codec according to the determination.

13. The baseband modem of claim 11, wherein the controller is further adapted to control activation of the speech recognition means according to the determination such that the speech recognition means is activated if the voice signal is a voice command and the speech recognition means is deactivated if the voice signal is voice communication.

14. The baseband modem of claim 13, wherein the controller is further adapted to power on registers of a feature vector extraction block and a speech recognition block if the voice signal is a voice command and power off registers of the feature vector extraction block and the speech recognition block if the voice signal is voice communication.

15. The baseband modem of claim 1, wherein the first sampling rate is optimized for speech recognition.

16. The baseband modem of claim 15, wherein the first sampling rate is in a range of approximately 12 kHz to approximately 32 kHz.

17. The baseband modem of claim 16, wherein the first sampling rate is approximately 16 kHz.

18. The baseband modem of claim 1, wherein the second sampling rate is optimized for voice communication.

19. The baseband modem of claim 18, wherein the second sampling rate is approximately 8 kHz.

20. The baseband modem of claim 1, wherein the audio codec is further adapted to perform pulse code modulation on the voice signal.

21. The baseband modem of claim 1, wherein the baseband modem is implemented in a mobile communication terminal.

22. A mobile communication terminal comprising:

an audio codec adapted to modulate a voice signal using one of a first sampling rate and a second sampling rate;

a feature vector extraction block adapted to extract at least one feature vector from the modulated voice signal;

a speech recognition block adapted to perform speech recognition using the at least one feature vector extracted by the feature vector extraction block; and

a vocoder adapted to vocode the modulated voice signal,

wherein the audio codec encodes the voice signal using the first sampling rate if the voice signal is a voice command and the audio codec encodes the voice signal using the second sampling rate if the voice signal is voice communication.

23. The mobile communication terminal of claim 22, further comprising a buffer adapted to store the encoded voice signal.

24. The mobile communication terminal of claim 23, wherein the buffer comprises a ping-pong buffer.

25. The mobile communication terminal of claim 22, further comprising a buffer adapted to store the at least one feature vector extracted from the modulated voice signal.

26. The mobile communication terminal of claim 22, further comprising a controller adapted to determine that the voice signal is one of a voice command and voice communication.

27. The mobile communication terminal of claim 26, wherein the controller is further adapted to determine that the voice signal is one of a voice command and voice communication according to a user selection.

28. The mobile communication terminal of claim 27, wherein the controller is further adapted to determine the sampling rate used by the audio codec according to the determination.

29. The mobile communication terminal of claim 27, wherein the controller is further adapted to control activation of the feature vector extraction block and the speech recognition block such that the feature vector extraction block and the speech recognition block are activated if the voice signal is a voice command and the feature vector extraction block and the speech recognition block are deactivated if the voice signal is voice communication.

30. The mobile communication terminal of claim 29, wherein the controller is further adapted to power on registers of the feature vector extraction block and the speech recognition block if the voice signal is a voice command and power off registers of the feature vector extraction block and the speech recognition block if the voice signal is voice communication.

31. The baseband modem of claim 22, wherein the first sampling rate is optimized for speech recognition.

32. The mobile communication terminal of claim 31, wherein the first sampling rate is in a range of approximately 12 kHz to approximately 32 kHz.

33. The mobile communication terminal of claim 31, wherein the first sampling rate is approximately 16 kHz.

34. The baseband modem of claim 22, wherein the second sampling rate is optimized for voice communication.

35. The mobile communication terminal of claim 34, wherein the second sampling rate is approximately 8 kHz.

36. The mobile communication terminal of claim 34, wherein the feature vector extraction block is implemented in hardware.

37. The mobile communication terminal of claim 22, wherein the feature vector extraction block is implemented in software.

38. The mobile communication terminal of claim 22, wherein the audio codec is further adapted to perform pulse code modulation on the voice signal.

39. A method of performing speech recognition and speech communication in a baseband modem, the method comprising:

determining that a voice signal is one of a voice command and voice communication; and

modulating the voice signal using a first sampling rate and performing speech recognition of the modulated voice signal if the voice signal is determined to be a voice command and modulating the voice signal using a second sampling rate and performing vocoding of the modulated voice signal if the voice signal is determined to be voice communication.

40. The method of claim 39, wherein performing speech recognition of the modulated voice signal comprises:

extracting at least one feature vector from the modulated voice signal; and

performing speech recognition using the at least one feature vector.

41. The method of claim 40, wherein performing speech recognition of the modulated voice signal further comprises:

storing the modulated voice signal in a buffer; and

extracting the at least one feature vector from data stored in the buffer.

42. The method of claim 40, wherein performing speech recognition of the modulated voice signal further comprises storing the at least one feature vector extracted from the modulated voice signal in a buffer.

43. The method of claim 40, further wherein extracting the at least one feature vector from the modulated voice signal is implemented in hardware.

44. The method of claim 40, further wherein extracting the at least one feature vector from the modulated voice signal is implemented in software.

45. The method of claim 39, further comprising determining that the voice signal is one of a voice command and voice communication according to a user selection.

46. The method of claim 39, further comprising controlling activation of a feature vector extraction block and a speech recognition block such that the feature vector extraction block and the speech recognition block are activated if the voice signal is a voice command and the feature vector extraction block and the speech recognition block are deactivated if the voice signal is voice communication.

47. The method of claim 46, further comprising powering on registers of the feature vector extraction block and the speech recognition block if the voice signal is a voice command and powering off registers of the feature vector extraction block and the speech recognition block if the voice signal is voice communication.

48. The method of claim 39, further comprising modulating the voice signal using a first sampling rate optimized for speech recognition.

49. The method of claim 48, further comprising modulating the voice signal using a first sampling rate in a range of approximately 12 kHz to approximately 32 kHz.

50. The method of claim 48, further comprising modulating the voice signal using a first sampling rate of approximately 16 kHz.

51. The method of claim 39, further comprising modulating the voice signal using a second sampling rate optimized for voice communication.

52. The method of claim 51, further comprising modulating the voice signal using a second sampling rate of approximately 8 kHz.

53. The method of claim 39, further comprising performing pulse code modulation on the voice signal.

54. The method of claim 39, wherein the baseband modem is implemented in a mobile communication terminal.