534.87
. . , . .

. . , . .
: , , , .
Abstract. The article considers an algorithm of isolating the pauses in a speech signal. The developed algorithm is based on the use of the theory of active perception adapted for the analysis of speech signals. Results of the experiments confirm the possibility of the offered algorithm application for solving the problem.
Key words: digital signal processing, theory of active perception, speech signal analysis, pauses isolation in a speech signal.

. , .
, , .. .
. , , . 0,1 [1]. ( ) 0,75 , 0,5 1,5 [2, 3].
, , 0,1 1,5 .
, - . . , [4]:
1) , [5];
2) ;
3) , .
[6, 7].
, . [8] , . .
- [9].
, [10, 11], , .
[12] , . , .
, , [13]. DSP- .
, - , [14].
, : , , .
/ (signal-to-noise ration, SNR) [15]:
5 - ; V - ; Ns N - .
- () . -ї :
1) ;
2) , ;
3) /- , ;
1.
2.
2.1.
4) i- , , 2-3.
, . , , .
, 0,1 , :
L = 0,1F, (1)
F - .
2.2.
. , . , . , , :
1) ft) L;
2) , ,
1 L
M = LI f (t) ;
L i=1
3) , i- , , ;
4) , ;
5) / , , ( / [16]).
, .
3.
Octave. AMD Turion 2 Dual-Core Mobile M500, 2,20 , 4 .
. 1. -16 , - 16 , - 8,192 . . 2. L 1024 (64 ). (1) (. 3) -
, . 128 (8 ). , , 1 [16].
. 1.
. 2.
(. 4) , ( ) 0,02, ( ) - 0,017. 0,18 .
:
1) ():
- ;
- ;
2) :
- ;
- ;
- ;
- .
. 3.
. 4.
. , , .
, , 128 . 0,14 .
, , . 1. 1 ( , ), 2 - ( , ).
() , , . 2.
1


- - -
1 2 1 2 1 2 1 2 1 2 1 2
30 0 6 0 6 0 0 5 5,6 0 5 0 5,8
20 0 6 0 6,7 3,1 3,3 4,3 5,9 0 6 0 4,2
10 0 7,1 10 3,8 30,6 14,6 3,7 3,9 1,5 3,5 5,3 3,3
5 1,5 7,8 32,4 1,9 35,4 40,2 3,8 4,4 14,8 3,5 20,1 1,5
0 31,4 4 51,8 0 35,5 46,7 4 2,4 40,1 3,7 20,7 3,6
2
( )

- - -
30 0,00171 0,00405 0,00260 0,00091 0,00089 0,00083
20 0,00559 0,01327 0,00813 0,00454 0,00445 0,00417
10 0,02021 0,04034 0,01951 0,01362 0,01423 0,00835
5 0,03495 0,05762 0,03903 0,02088 0,02313 0,01669
0 0,05013 0,11537 0,06992 0,84024 0,03503 0,03004
. 5-7 , , .
:
1) , - 2-3 ;
2) 2 (1 - 1; 2 - 2; 1 << 2), 1 , 2, ;
3) 1 2 (1 - 1; 2 - 2; 1 << 2), 1 , 2.
, ї [15], , .

. , , , - . , , .
0.4 -
0 2
)
)
10
10
12
10
10
. 5. , ( = 10 ): - ; -
)
10
. 6. , ( = 0 ): - ; - (. . 92)
0,4 -
0.2
)
. 6.
10
12
10
10 12
10
)
10
1| 1 1 1 1 1 1 0 2 4 6 8 10 12 14 4 10 )
| 0 2 4 1 6 1 1 1 8 10 12
- ^ 10 1| .1^1
12
10
. 7. , ( = 20 ): - ; -
.
: , , .
, .

1. , . . / . . , . . // . - 1955. - . 3. - . 5-17.
2. , . . / . . - // . - .-. : , 1966. - . 31-44.
3. Goldman-Eister, F. Pauses, clauses, centences / F. Goldman-Eister // Language and Speech. - 1972. - V. 15, 3. - P. 103-113.
4. , . . / . . , . . , . . , . . , . . // . - 2005. - . 36, 1. - . 3-23.
5. , . / . , . , , . , . ; . . . . , . . , . . , . . , . . . - . : , 2003. - 672 .
6. , . . : / . . . - . : - , 2001. - 234 .
7. , . . . : / . . . - . : - , 2004. - 221 .
8. Beritelli, F. A robust voice activity detector for wireless communications using soft computing / F. Beritelli, S. Casale, A. Cavallaro // IEEE Journal on Selected Areas in Communications. - 1998. - V. 16, 9. - P. 1818-1829.
9. McKinley, B. L. Model based speech pause detection / B. L. McKinley,
G. H. Whipple // Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 97). - Munich, Germany, 1997. - V. 2. -P. 1179-1182.
10. Sohn, J. A statistical model-based voice activity detection / J. Sohn, N. S. Kim, W. Song // IEEE Signal Processing Letters. - 1999. - V. 6, 1. - P. 1-3.
11. Cho, Y. D. Analysis and improvement of a statistical model-based voice activity detector / Y. D. Cho, A. Kondoz // IEEE Signal Processing Letters. - 2001. - V. 8, 10. -P. 276-278.
12. Gazor, S. A soft voice activity detector based on a Laplacian-Gaussian model / S. Gazor, W. Zhang // IEEE Transactions on Speech and Audio Processing. - 2003. -V. 11, 5. - P. 498-505.
13. Sheikhzadeh, H. Real-time implementation of HMM-based MMSE algorithm for speech enhancement in hearing aid applications / H. Sheikhzadeh, R. L. Brennan,
H. Sameti // Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 95). - Detroit, Mich, USA, 1995. - V. 1. -P. 808-811.
14. Rezayee, A. An adaptive KLT approach for speech enhancement / A. Rezayee, S. Gazor // IEEE Transactions on Speech and Audio Processing. - 2001. - V. 9, 2. -P. 87-95.
15. Pwint, M. Speech / Nonspeech Detection Using Minimal Walsh Basis Functions / M. Pwint, F. Sattar // EURASIP Journal on Audio, Speech, and Music Processing. -2006. - V. 2007. - P. 3-12.
16. / . . . . - . : . - , 1978. - 360 .

, , , . . .
E-mail: iamuser@inbox.ru
, , , . . .
E-mail: utrobin-va@yandex.ru
Gai Vasily Evgenyevich
Candidate of engineering sciences, associate
professor, sub-department of computing
systems and technologies, Nizhny
Novgorod State University named
after R. E. Alekseev
Utrobin Vladimir Alexandrovich Doctor of engineering sciences, professor, sub-department of computing systems and technologies, Nizhny Novgorod State University named after R. E. Alekseev
534.87 , . .
/ . . , . . // . . . - 2011. - 4 (20). - . 85-94.