Method and apparatus for improving disordered voice

• US 9,646,602 B2
• Filed: 06/20/2014
• Issued: 05/09/2017
• Est. Priority Date: 06/21/2013
• Status: Active Grant

• Current Assignee: SNU R&DB Foundation (Seoul National University)
• Original Assignee: SNU R&DB Foundation (Seoul National University)
• Inventors: Kim, Hee Jin, Kim, Dong Wook, Sung, Myung Whun, Kwon, Tack Kyun, Kim, Wook Eun, Kim, Woo Il, Sung, Mee Young
• Primary Examiner(s): Desir, Pierre-Louis
• Assistant Examiner(s): KIM, JONATHAN C
• Application Number:
  US14/311,115
• Publication Number:
  US 20140379348A1
• Time in Patent Office: 1,054 Days
• Field of Search: None
• CPC Class Codes: G10L 13/04 Details of speech synthesis systems, e.g. synthesiser structure or memory management
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 2015/025 Phonemes, fenemes or fenones being the recognition units
• CPC Class Codes: G10L 25/15 the extracted parameters being formant information
• CPC Class Codes: G10L 25/27 characterised by the analysis technique
• CPC Class Codes: G10L 25/60 for measuring the quality of voice signals
• CPC Class Codes: G10L 25/75 for modelling vocal tract parameters

Abstract

There is provided a method and an apparatus for processing a disordered voice. A method for processing a disordered voice according to an exemplary embodiment of the present invention includes: receiving a voice signal; recognizing the voice signal by phoneme; extracting multiple voice components from the voice signal; acquiring restored voice components by processing at least some disordered voice components of the multiple voice components by phoneme; and synthesizing a restored voice signal based on at least the restored voice components.

Claims

1. A disordered voice processing method comprising:

receiving a nasal sound signal and an oral sound signal, and the nasal sound signal and the oral sound signal are acquired from an input unit with a separation means for separating the oral sound signal and the nasal sound signal from each other;

recognizing the nasal sound signal and the oral sound signal by phoneme;

classifying each of the recognized nasal sound signal and oral sound signal into a normal voice signal and a disordered voice signal using a phoneme characteristic table, wherein the phoneme characteristic table includes information of nasal sound and oral sound of normal voice separately;

extracting multiple voice components from each of the nasal sound signal and the oral sound signal;

acquiring restored voice components by processing at least a portion of disordered voice components of the multiple voice components on a basis of a phoneme, wherein the restored voice components are acquired for each of the nasal sound signal and the oral sound signal using the phoneme characteristic table; and

synthesizing a restored nasal sound signal and a restored oral sound signal based on at least the restored voice components to generate a restored voice signal.

2. The disordered voice processing method according to claim 1,

wherein the multiple voice components include vocal tract components and glottal components, and

the disordered voice components are vocal tract components.

3. The disordered voice processing method according to claim 1,

wherein the acquiring the restored voice components includes acquiring the restored voice component by processing the disordered voice components with respect to a disordered voice signal on a basis of a phoneme.

4. The disordered voice processing method according to claim 1,

wherein the classifying each of the recognized nasal sound signal and oral sound signal includes classifying the disordered voice signal by applying heuristics.

5. The disordered voice processing method according to claim 1,

wherein the classifying each of the recognized nasal sound signal and oral sound signal includes determining whether or not each of the nasal sound signal and the oral sound signal is disordered on a basis of a phoneme.

6. The disordered voice processing method according to claim 1,

wherein the classifying each of the recognized nasal sound signal and oral sound signal is based on the extracted multiple voice components.

7. The disordered voice processing method according to claim 1, further comprising:

adjusting amplitude or a length of the disordered voice signal.

8. The disordered voice processing method according to claim 1,

wherein the disordered voice component is a disordered voice component associated with a consonant, and

acquiring restored voice components includes acquiring the restored voice components by adjusting the disordered voice component with respect to at least one of aspiration, tense and turbulence.

9. The disordered voice processing method according to claim 1,

wherein the disordered voice components are disordered voice components with respect to a consonant, a vowel, or a combination of a consonant and a vowel.

10. The disordered voice processing method according to claim 1,

wherein synthesizing the restored nasal sound signal and the restored oral sound signal includes the use of Linear Predictive Coding (LPC).

11. The disordered voice processing method according to claim 1,

wherein the extracting includes extracting the multiple voice components using Linear Predictive Coding (LPC).

12. The disordered voice processing method according to claim 1,

wherein generating the restored voice components includes generating the restored voice components based on a phoneme characteristic table.

13. The disordered voice processing method according to claim 1, further comprising:

filtering the restored nasal sound signal and the restored oral sound signal to compensate the restored voice components.

14. The disordered voice processing method according to claim 1, further comprising:

filtering the restored nasal sound signal and the restored oral sound signal through a lip radiation filter.

15. A disordered voice processing apparatus comprising:

an input unit configured to input a nasal sound signal and an oral sound signal including a separation means for separating the oral sound signal and the nasal sound signal from each other;

a memory configured to store the nasal sound signal and the oral sound signal; and

a processor configured to generate a restored voice signal from the nasal sound signal and the oral sound signal,

wherein the processor is configured;

to recognize the nasal sound signal and the oral sound signal on a basis of a phoneme;

classify each of the recognized nasal sound signal and oral sound signal into a normal voice signal and a disordered voice signal using a phoneme characteristic table, wherein the phoneme characteristic table includes information of nasal sound and oral sound of normal voice separately;

to extract multiple voice components from each of the nasal sound signal and the oral sound signal;

to acquire restored voice components by processing at least a portion of disordered voice components of the multiple voice components on a basis of a phoneme, wherein the restored voice components are acquired for each of the nasal sound signal and the oral sound signal using the phoneme characteristic table; and

to synthesize a restored nasal sound signal and a restored oral sound signal based on at least the restored voice components to generate the restored voice signal.

16. The disordered voice processing apparatus according to claim 15,

wherein the multiple voice components include vocal tract components and glottal components, and

the disordered voice components are vocal tract components.

17. The disordered voice processing apparatus according to claim 15,

wherein the processor is further configured to adjust amplitude or a length of the disordered voice signal.