Speech recognition system and method for generating a mask of the system

• US 8,392,185 B2
• Filed: 08/19/2009
• Issued: 03/05/2013
• Est. Priority Date: 08/20/2008
• Status: Active Grant

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: Nakadai, Kazuhiro, Takahashi, Toru, Okuno, Hiroshi
• Primary Examiner(s): GUERRA-ERAZO, EDGAR X
• Application Number:
  US12/543,759
• Publication Number:
  US 20100082340A1
• Time in Patent Office: 1,294 Days
• Field of Search: 704/243, 704/233, 704/234, 704/231, 704/256, 704/255, 704/236, 704/240, 704/239, 704/270, 704/270.1, 704/275, 704/235, 704/246
• US Class Current: 704/233
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 21/0272 Voice signal separating

Abstract

The speech recognition system of the present invention includes: a sound source separating section which separates mixed speeches from multiple sound sources; a mask generating section which generates a soft mask which can take continuous values between 0 and 1 for each separated speech according to reliability of separation in separating operation of the sound source separating section; and a speech recognizing section which recognizes speeches separated by the sound source separating section using soft masks generated by the mask generating section.

Claims

1. A speech recognition system comprising:

multiple sound sources;

a sound source separating section which separates mixed speeches from the multiple sound sources; and

at least one processor configured to;

generate a soft mask which can take continuous values between 0 and 1 for each separated speech according to reliability of separation in separating operation of the sound source separating section, and

recognize speeches separated by the sound source separating section using the soft masks,

wherein the reliability of separation R(f,t) is defined as

2. A speech recognition system according to claim 1, wherein the soft masks are determined using a sigmoid function

1/(1+exp(−

a(R−

b))

where R represents reliability of separation and a and b represent constants.

3. A speech recognition system according to claim 1, wherein the soft masks are determined using a probability density function of a normal distribution, which has a variable R which represents reliability of separation.

10. A speech recognition system according to claim 1, wherein assuming that

μ

1 and μ

2 (μ

1<

μ

2)

indicate mean values,

σ

1 and σ

2

indicate standard deviations, and R indicates the reliability of separation, the mean values and standard deviations

μ

1, μ

2, σ

1 and σ

2 are estimated by fitting a histogram of the reliability of separation R with a first probability density function of normal distribution f1(R) which has

(μ

1,σ

1)

and a second probability density function of normal distribution f2(R) which has

(μ

2,σ

2)

and the soft mask is generated using f1(R), f2(R),

μ

1 and μ

2.

4. A method for generating a soft mask for a speech recognition system, the method comprising:

separating, at a sound source separating section of the speech recognition system, mixed speeches from multiple sound sources;

generating, at a mask generating section of the speech recognition system, a soft mask which can take continuous values between 0 and 1 for each separated speech according to reliability of separation in separating operation of the sound source separating section;

recognizing, at a speech recognizing section of the speech recognition system, speeches separated by the sound source separating section using soft masks generated by the mask generating section, the soft mask being determined using a function of the reliability of separation, which has at least one parameter;

determining a search space of said at least one parameter;

obtaining a speech recognition rate of the speech recognition system while changing a value of the speech recognition system in the search space; and

setting the value which maximizes a speech recognition rate of the speech recognition system to said at least one parameter,

wherein the reliability of separation R(f,t) is defined as

5. A method for generating a soft mask for a speech recognition system, the method comprising:

separating, at a sound source separating section of the speech recognition system, mixed speeches from multiple sound sources;

generating, at a mask generating section of the speech recognition system, a soft mask which can take continuous values between 0 and 1 for each separated speech according to reliability of separation in separating operation of the sound source separating section;

recognizing, at a speech recognizing section of the speech recognition system, speeches separated by the sound source separating section using soft masks generated by the mask generating section, the soft mask being determined using a function of the reliability of separation, which has at least one parameter;

obtaining a histogram of the reliability of separation; and

determining a value of said at least one parameter from a form of the histogram of the reliability of separation,

wherein the reliability of separation R(f,t) is defined as

6. A method for generating a soft mask for a speech recognition system according to claim 5, wherein assuming that

μ

1 and μ

2 (μ

1<

μ

2)

indicate mean values and

σ

1 and σ

2

indicate standard deviations and R indicates reliability of separation, the mean values and standard deviations

μ

1, μ

2, σ

1 and σ

2

are estimated by fitting the histogram of reliability of separation R with a first probability density function of normal distribution f1(R) which has

(μ

1,σ

1)

and a second probability density function of normal distribution f2(R) which has

(μ

2,σ

2)

and the soft mask is generated using f1(R), f2(R),

μ

1 and μ

2.

7. A method for generating a soft mask for a speech recognition system according to claim 6, wherein assuming that a value of the soft mask is S(R) and f(R)=f1(R)+f2(R),

S(R)=0 when R<

μ

1,

S(R)=f2(R)/f(R) when μ

1≦

R≦

μ

2

S(R)=1 when μ

2<

R.

8. A method for generating a soft mask for a speech recognition system according to claim 6, wherein assuming that a value of the soft mask is S(R),

9. A method for generating a soft mask for a speech recognition system according to claim 6, wherein a value of R at the intersection of f1(R) and f2(R) which satisfies

μ

1<

R<

μ

2

is set to b and a is determined such that

1/(1+exp(−

a(R−

b))

is fit to

f2(R)/f(R)

and the value of the missing feature mask (MFM) S(R) is determined by

S(R)=1/(1+exp(−

a(R−

b)).