MOBILE RADIO TERMINAL, SPEECH CONVERSION METHOD AND PROGRAM FOR THE SAME

• US 20090083038A1
• Filed: 09/17/2008
• Published: 03/26/2009
• Est. Priority Date: 09/21/2007
• Status: Active Grant

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Imoto, Kazunori
• Granted Patent: US 8,209,167 B2
• Time in Patent Office: Days
• US Class Current: 704/270
• CPC Class Codes: G10L 2021/0135 Voice conversion or morphing
• CPC Class Codes: H04M 1/60 including speech amplifiers
• CPC Class Codes: H04M 1/72448 with means for adapting the functionality of the device according to specific conditions

Abstract

The mobile radio terminal includes a speech input unit which inputs a speech signal obtained from speech of a speaking person, an estimating unit which estimates a speech style of the speaking person from the speech signal, and a converting unit which converts the speech signal into a converted speech signal in accordance with the estimated speech style.

Claims

1. A mobile radio terminal, comprising:

a speech input unit which inputs a speech signal obtained from speech of a speaking person;

an estimating unit which estimates a speech style of the speaking person from the speech signal; and

a converting unit which converts the speech signal into a converted speech signal in accordance with the estimated speech style.

2. The terminal according to claim 1, wherein the estimating unit comprises a detecting unit which detects a feature quantity indicating a speech feature from the speech signal, and a style estimating unit which estimates the speech style in accordance with the feature quantity.

3. The terminal according to claim 2, wherein the converting unit converts the speech signal into the converted speech signal in accordance with the feature quantity detected by the detecting unit.

4. The terminal according to claim 2, further comprising an encoding unit which encodes the converted speech signal converted by the converting unit, in accordance with the feature quantity detected by the detecting unit.

5. The terminal according to claim 1, wherein the estimating unit comprises:

a detecting unit which detects a signal intensity from the speech signal; and

a style estimating unit which estimates the speech style in accordance with the signal intensity.

6. The terminal according to claim 1, wherein the estimating unit comprises:

a detecting unit which detects an intensity of a pitch component from the speech signal; and

a style estimating unit which estimates the speech style in accordance with the intensity of the pitch component.

7. The terminal according to claim 1, wherein the estimating unit comprises:

a detecting unit which detects each of intensities of spectrum components in two frequency bands from the speech signal; and

a style estimating unit which estimates the speech style in accordance with a difference between the intensities of the spectrum components in the two frequency bands.

8. The terminal according to claim 1, wherein the estimating unit comprises:

a storing unit which stores dictionary data preliminarily associating a feature quantity with a speech style;

a detecting unit which detects the feature quantity indicating a speech feature from the speech signal; and

a style estimating unit which estimates the speech style in accordance with the detected feature quantity, by referring to the dictionary data.

9. The terminal according to claim 1, further comprising an accepting unit which accepts setting of an operation mode,

wherein the estimating unit estimates the speech style in accordance with the set operation mode.

10. The terminal according to claim 9, further comprising a detecting unit which detects a feature quantity indicating a speech feature from the speech signal,

wherein the estimating unit estimates the speech style in accordance with the set operation mode and the feature quantity.

11. The terminal according to claim 10, wherein the estimating unit estimates the speech style in accordance with the set operation mode and, if the estimated speech style is not a preset speech style, estimates the speech style in accordance with the feature quantity.

12. The terminal according to claim 1, wherein the converting unit executes conversion by adding a pitch component to the speech signal.

13. The terminal according to claim 1, wherein the converting unit comprises:

a detecting unit which detects a feature quantity indicating a speech feature from the speech signal;

a storing unit which stores the feature quantity and conversion data in association with each other; and

a speech converting unit which converts the speech signal into the converted speech signal by using the conversion data corresponding to the detected feature quantity.

14. The terminal according to claim 13, further comprising an encoding unit which encodes the converted speech signal by using the detected feature quantity.

15. The terminal according to claim 13, further comprising:

a learning detecting unit which detects the feature quantity indicating the speech feature from the speech signal;

a generating unit which generates the conversion data in accordance with the feature quantity detected by the learning detecting unit; and

a recording unit which records the feature quantity detected by the learning detecting unit and the conversion data generated by the generating unit, in association with each other, in the storing unit.

16. The terminal according to claim 15, further comprising an encoding unit which encodes the converted speech signal by using the feature quantity detected by the learning detecting unit.

17. The terminal according to claim 15, wherein the estimating unit estimates the speech style in accordance with the feature quantity detected by the learning detecting unit.

18. The terminal according to claim 15, wherein the learning detecting unit is the detecting unit.

19. The terminal according to claim 15, wherein if the speech style estimated by the estimating unit is a preset speech style, the recording unit records the feature quantity detected by the learning detecting unit and the conversion data generated by the generating unit, in association with each other, in the storing unit.

20. A method of converting speech, comprising:

inputting a speech signal obtained from speech of a speaking person;

estimating a speech style of a speaking person from the speech signal; and

a converting the speech signal into a converted speech signal in accordance with the estimated speech style.

21. A computer readable storage medium storing instructions of a computer program which when executed by a computer results in performance of steps comprising:

inputting a speech signal obtained from speech of a speaking person;

estimating a speech style of a speaking person from the speech signal; and

a converting the speech signal into a converted speech signal in accordance with the estimated speech style.