Acoustic Representation of BODO and RABHA Phonemes by warse1

VIEWS: 24 PAGES: 9

									                                                                                                                                 ISSN No.
                                                                        Volume 1, No.1, July – August
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9   2012
                                      International Journal of Computing, Communications and Networking
                                                   Available Online at http://warse.org/pdfs/ijccn01112012.pdf

                              Acoustic Representation of BODO and RABHA Phonemes

                                                        Jyotismita Talukdar1, Nabankur Pathak2
                               1
                                   Asia Institute of Technology, Bangkok, Thailand, E-mail:jyotismita4@gmail.com
                                               2
                                                 Gauhati University, India, E-mail:phtassam@gmail.com



ABSTRACT                                                                                      formant frequency is maximum in case of isolated vowels,
                                                                                              but when the vowels are placed in the nucleus of a structure
In this paper we studied the spectral features of Bodo and                                    like CV, VC or CVC, the formant frequency decreases.
Rabha Phonemes. The spectral features are studied using
formant frequency and Cepstral coefficients. Depending on                                     Keywords: Acoustic Representation, Phonemes, Cepstral
the analysis on cepstral features and formant frequencies of                                  Features
Bodo and Rabha phonemes and words we observed that
significant variation of cepstral coefficients are observed                                   1. INTRODUCTION
among the Bodo vowels. The cepstral variation is found to
be maximum with respect to vowel /o/ and minimum                                              The Bodos and the Rabhas are the early ethnic and linguistic
corresponding to vowel /u/, in case of male speakers.                                         communities settled in the North-Eastern part of India. The
Similarly, for female Bodo speakers, the maximum variation
                                                                                              Bodos belong to a larger group of ethnicity called the Bodo-
of cepstral measure is found corresponding to vowels /o/ and
                                                                                              Kachari. Racially, they belong to a Mongoloid stock of the
minimum in case of /i/.In case of Rabha vowels, i.e., /o/, /a/,
                                                                                              Indo-Mongoloids or         Indo-Tibetans.    Mythologically,
/i/, ./e/,, /u/ and /w/ for both male and female speakers the
range of variation of the cepstral coefficient is found to be                                 according to Dr. Suniti Kumar Chatterjee, a well-known
maximum in case of male speakers with respect to vowel /u/                                    historian, the Bodos are “the Offspring of son of Vishnu and
and minimum with respect to vowel /o/. In case of female                                      mother Earth”, who are termed as Kiratas during the epic
speaker, the maximum variation of cepstral co-efficient is                                    period. They are recognized as a plain tribe in the 6th
found in case of vowel /o/ and minimum with respect to                                        schedule of the constitution of India. Historically, there are
vowel /e/. This observation may be helpful in sex                                             different views on the early migration of the Mongolian into
determination for both Bodo and Rabha speakers.The range                                      the North-Eastern part of India. Some of them are:
of variation of cepstral coefficients for Bodo and Rabha male
is found within the range of 3.8177 >CBodo>1.1523 and                                         According to Grierson’s “The Linguistic Survey of India”,
8.1329>CRabha>2.0579 respectively. The range of variation                                     the Mongolian settled in old Assam, migrated from Hoang-
for female is found 1.9578>CBodo>0.9276 and                                                   Ho and Yangtze River banks and scattered and dwelt in
7.6546>CRabha>2.4127. i.e. the variation of cepstral                                          different river banks of the state. The upper course of the
features for Bodo vowels is less (Male-2.6654; Female-                                        Yangtz and Hoang-Ho in the North-West China were the
1.0302) with respect to the Rabha vowels (Male-6.0750;                                        original home of the Tibeto-Burman races. The hierarchy of
Female-5.2419) i.e., the former is stable as compared to the                                  Bodo community is shown in figure .
latter. The investigation have shown that the range of




                                                                    Hierarchy of Bodo & Rabha Languages
                                                                                                                                                           1


@ 2012, IJCCN All Rights Reserved
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9


Speech Data Collection for Acoustic Representation                                               2. LPC ANALYSIS

Typically, the spoken language data can be classified                                            Linear prediction is a method for signal source modelling
based on                                                                                         dominant in speech signal processing and having wide
   Mode of speech                                                                               application in other areas. Linear Predictive Coding (LPC) is
   Medium of recoding                                                                           one of the most powerful speech analysis techniques. The
   Language Dialects                                                                            glottis (the space between the vocal cords) produces the
   Environment                                                                                  sound, which is characterized by its intensity (loudness) and
                                                                                                 frequency (pitch). The vocal tract (the throat, the mouth and
                                                                                                 the nasal cavity) forms the tube, which is characterized by its
In the present study, speech data is collected from the native                                   resonance frequencies, which are called formants.
speakers of Bodo and Rabha language who are fluent in
speaking and writing the language. Male and female speaker
of age between 15 to 30 years, possessing a pleasant and a                                       The basic problem of the LPC system is to determine the
good voice quality are chosen to record the data. The                                            formants from the speech signal. The solution of this
recording is done one-by one manner. The speakers were                                           problem is a difference equation, which expresses each
instructed to read each word or sentence naturally, without                                      sample of the signal as a linear combination of previous
emotions and expression. They were asked to speak clearly                                        samples. Such an equation is called a linear predictor i.e.
and to keep their normal speaking rate and volume. To keep                                       Linear Predictive Coding. The coefficients of the difference
the recording consistent, both in phonetic and prosodic                                          equation (the prediction coefficients) characterize the
(within the framework of symbolic Prosody) terms, an                                             formants. Therefore, the LPC system needs to estimate these
expert in acoustic phonetics supervised the recording. The                                       coefficients. The estimation is made by minimizing the mean
average duration of recording session was about 4 hours (3                                       square error between the predicted signal and the actual
recording session) for each speaker (Male & Female).                                             signal.
We have recorded the following data sets for analysis of the
cepstral coefficients of vowel phonemes and formant                                              The basic idea behind the LPC model is that a given speech
frequencies of some selected Bodo and Rabha words.                                               sample         at time n, can be approximated as a linear
 Bodo and Rabha vowel phonemes for cepstral analysis.                                           combination of the past p speech samples (Rabiner & Juang,
 Selected word sets of V, CV, VC and CVC structure in                                           1993) such that (1)
   both languages for formant analysis.
                                                                                                                                                        (1)
The recording is done in audio editing software Cool Edit
Pro and the analysis was done in MATLAB 7.1. Each
                                                                                                 Where the coefficients are 1 2
                                                                                                                                 a , a ,...a
                                                                                                                                         n assumed to be
digitized voice uttered, is divided or blocked into 50 frames                                    constants over the speech analysis frame. The equation (1)
of duration 20 millisecond (ms). Every frame contains 441                                        can be converted to an equality by including an excitation
samples and for each frame 20 cepstral coefficients have                                         term Gu(n),
been calculated. The spectral characteristics of six Bodo and
Rabha vowels, corresponding to male and female speakers
were investigated. Approximately 12 samples were averaged
to obtain one coefficient. Firstly, 10th frame of all                                                                                                   (2)
utterances of male and female speakers have been
                                                                                                 Where          normalized excitation and G is the gain of
considered for analysis. The variation of the cepstral
                                                                                                 excitation. Expressing equation (2) in Z domain we get the
coefficients for the Bodo and Rabha vowels corresponding                                         relation:
to the selected speakers have been shown in Table-(1) &
Table-(2) and depicted in Figures-(3 & 4) and Figures-(6 &
7). However, from continuous frame wise analysis, it is
observed that: 2, 4, 6, and 8 frames for Bodo speaker                                                                                                   (3)
(Figure-5) and 9, 14, 16 and 17 frames for Rabha speaker                                         Leading to the transfer function:
(Figure-8) have shown distinct variation of the cepstral
coefficients for male and female speakers.
                                                                                                                                                      (4)
                                                                                                 based on our knowledge that the actual excitation function
                                                                                                 for speech is essentially either voiced speech sounds or an
                                                                                                 unvoiced sound.
                                                                                                                                                               2


@ 2012, IJCCN All Rights Reserved
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9



The relation between       and         is defined (based on                                                       s (m  i) s (m  k )
                                                                                                                          n   n
the speech production model Figure-1.1)                                                          This term m                                are related to the short
                                                                                                 term covariance of sn(m) i.e.,

                                                                                                                                                            (15)
                                                              (5)
We consider the linear combination of past speech samples                                        Which can be expressed in compact notation as,
as the estimate               , defined as,

                                                                                                                                                       (16)
                                                                                                 Which describe a set of p equations. It is readily shown that
                                                              (6)                                the minimum mean-square error,             , can be expressed as :
The predictor error,               , is defined as ,
                                                                                                                                                            (17)

                                                                                                 thus the minimum mean-squared error consists of a fixed
                                                              (7)                                term          and is depend on the predictor coefficients.
And the error transfer function is,
                                                                                                 To solve Equation (16) for the optimum coefficients               ,
                                                                                                 we have to compute                        , for                and
                                  =1-                                      (8)
                                                                                                              , and then solve the resulting set of p
                                                                                                 simultaneous equations. A method to solve these equations
The basic problem of linear prediction analysis is to                                            and compute the coefficients is the autocorrelation method.
determine the set of predictor coefficient    , directly from
the speech signal so that the speech properties of the digital                                   The LPC-Cepstral Co-efficient
filter match those of the speech waveform within the
analysis window.
                                                                                                 In the present study, LPC-based cepstral coefficients and
                                                                                                 phonetically important parameters are used as feature
To set up the equations that must be solved to determine the                                     vectors. Cepstral weighted feature vector is obtained for each
predictor coefficients, we define the short-term speech and                                      frame by block processing of continuous speech signals. The
error segments at time n as,                                                                     analog speech waveform is then sampled and quantized
                                                     (9)                                         analog-to-digital converter. To spectrally flatten the signal,
                                                     (10)                                        the speech signal has been subjected to the pre-emphasis
                                                                                                 procedure through a first order digital filter whose transfer
and tried to minimize the mean square error signal at time n,                                    function has been given by
                                  (11) Using equation (9)                                                                         , with                    (19)
& (10) we can write
                                                                                                 Consecutive speech signal are taken as a single frame. To
                                                                                                 reduce the undesired effect of Gibbs phenomenon, the
                                                                           (12)                  frames are multiplied by a windows function (Hamming
To solve the equation (4.12) we put                                                              window), which is given by (Proakis, & Manolakis,
                                                                                                 2004;Talukdar , P.H, 2010)

                                                                           (13)
giving
                                                                                                 Where N is the number of sample in a block. Now, each
                                                                           (14)                  frame of the windowed signal is next auto correlated to give

                                                                                                                                                            (20)

                                                                                                                                                                   3


@ 2012, IJCCN All Rights Reserved
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9


m=0, 1, 2…p
Where the highest auto correlated value                                               is the order of the                                                                                                   (21)
LPC analysis.
                                                                                                                                                                                 (22)
a. LPC Parameter Conversion to Cepstral Coeffecients                                                                                    Equation (4.30) shows the computation of cepstral
                                                                                                                                        coefficients C p+1, C p+2…C p.
The LPC cepstral coefficients, which are a set of values that                                                                           Generally,                       is taken for cepstral representation.
have been found to be more robust, reliable feature set for
speech recognition than the LPC coefficients. These
coefficients are obtained recursively as follows.

                       Where                                 is the gain term in the LPC
model.

                                                   Table 1: Range of variation of the cepstral coefficients corresponding to the male and female Bodo speaker

                                                                                                           Cepstral Coefficients
                                                                                        Male                                                            Female
                                                       Vowel         Max.              Min.      Range of variation                      Max      Min        Range of variation
                                                       /o/           2.2237            -1.5940   3.8177                                  1.9492   -0.0086    1.9578
                                                       /a/           1.6260            -0.9615   2.5875                                  0.9492   -0.0641               1.0133
                                                       /i/           1.1528            -0.1253   1.2781                                  0.9059   -0.0217               0.9276
                                                       /e/           1.2355            -0.6532   1.8887                                  0.9847   -0.0578               1.0425
                                                       /u/           1.0922            -0.0601   1.1523                                  1.1385   0.0690                1.2075
                                                       /w/           1.1832            -0.1541   1.3373                                  1.1843   -0.1674               1.3517



                          Figure 1. Cepstral characteristics of Bodo vowels for                                        Figure 2. Cepstral characteristics of Bodo vowels for
                          male speaker                                                                                 female speaker

                                             10                                       10                                                4                                   4

                                              0                                        0                                                2               /o/                 2            /a/

                                             -10                                     -10                                                0                                   0
                                                                   /o/                               /a/
                                             -20                                     -20                                                -2                                 -2
                                                   0           5             10            0     5            10                             0    5               10            0    5         10
                                              2                                        4                                                2                                   4
                                                                                                                        Amplitude(dB)
                             Amplitude(dB)




                                              1                    /i/                 2                                                1               /i/                 2            /e/
                                                                                                     /e/
                                              0                                        0                                                0                                   0

                                              -1                                      -2                                                -1                                 -2
                                                   0           5             10            0     5            10                             0    5               10            0    5         10
                                              2                                        2                                                4                                   4

                                              1                    /u/                 1                                                2             /u/                   2            /w/
                                                                                                     /w/
                                              0                                        0                                                0                                   0

                                              -1                                      -1                                                -2                                -2
                                                   0           5             10            0     5            10                             0    5              10           0      5         10
                                                                         Cepstral Coefficient                                                                 Cepstral Coefficient




                                                                                                                                                                                                                   4


@ 2012, IJCCN All Rights Reserved
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9

                                                                                                        Frame no.:2                                                                                                                       Frame no.:4
                                                                 1.4                                                                                                                            2
                                                                                                                                          Male                                                                                                                                 Male
                                                                 1.2                                                                      Female                                                                                                                               Female
                                                                                                                                                                                  1.5
                                                                  1

                                                                 0.8                                                                                                                            1




                                                                                                                                                                 agnitude(dB)
                               agnitude(dB)
                                                                 0.6
                                                                                                                                                                                  0.5
                                                                 0.4




                                                                                                                                                             LogM
                                                                                                                                                                                                0
                           LogM                                  0.2

                                                                  0                                                                                                             -0.5

                                                             -0.2
                                                                                                                                                                                          -1
                                                             -0.4

                                                             -0.6                                                                                                               -1.5
                                                                       0   2   4       6           8      10      12      14    16        18       20                                               0           2   4       6         8      10      12          14   16       18         20
                                                                                                 Cepstral Coefficients                                                                                                              Cepstral Coefficients

                                                                                                    Frame no.:16                                                                                                                          Frame no.:6
                                                                  8                                                                                                               1.5
                                                                                                                                          Male                                                                                                                                 Male
                                                                  7                                                                       Female                                                                                                                               Female

                                                                  6                                                                                                                             1

                                                                  5




                                                                                                                                                             LogMagnitude(dB)
                                              LogMagnitude(dB)




                                                                  4                                                                                                               0.5

                                                                  3

                                                                  2                                                                                                                             0

                                                                  1

                                                                  0                                                                                                             -0.5

                                                                  -1

                                                                  -2                                                                                                                      -1
                                                                       0   2   4       6           8      10      12      14    16        18       20                                               0           2   4       6         8      10      12          14   16       18         20
                                                                                                 Cepstral Coefficients                                                                                                              Cepstral Coefficients


                                                                               Figure 3. Distinction between Bodo Male & Female speaker in frame no 2,4,16 & 8


                                                                 Table 2: Range of variation of the Cepstral coefficients corresponding to the Male and Female Bodo speaker
                                                                                                             Cepstral Coefficients
                                                                                             Male                                               Female
                                                                     vowel    Max.        Min.         Range of variation Max.           Min.         Range of variation
                                                                     /o/      1.0057      -1.0522      2.0579                  3.9045    -3.7501      7.6546
                                                                     /a/      1.4964      -1.8083      3.3047                  2.0135    -2.0784      4.0919
                                                                     /i/      1.4086      -1.8085      3.2171                  1.9864    -1.9832      3.9696
                                                                     /e/      2.1054      -2.2054      4.3108                  0.9164    -1.4963      2.4127
                                                                     /u/      3.4942      -4.6387      8.1329                  1.0839    -1.6952      2.7791
                                                                     /w/      2.4834      -1.0627      3.5461                  1.7201    -0.8801      2.6002


                                              Figure 4. Distinction between the male and female Rabha speaker in frame no: 9,14,16 & 17
                                     Figure 5. Cepstral characteristics of Rabha vowels for     Figure 6. Cepstral characteristics of Rabha vowels
                                                         male speaker                                           for female speaker.
                                                                 5                                                5                                                                                     5                                               5

                                                                                                  /o/                          /a/                                                                                                /o/                                               /a/
                                                                 0                                                0                                                                                     0                                               0


                                                                 -5                                              -5                                                                                 -5                                                  -5
                                                                      0            5                     10           0               5                 10                                                  0           5                     10             0             5                     10
                                                                 2                                                2                                                                                 2 0                 5                     10        5
                                                                                                                                                                                Amplitude(dB)
                               Amplitude(dB)




                                                                                                                                                                                                                                                                                    /e/
                                                                                       /i/                                      /e/                                                                                         /i/
                                                                 0                                                0                                                                                 0                                                   0


                                                                 -2                                              -2                                                                             -2                                                      -5
                                                                      0            5                     10           0               5                 10                                                                                                   0             5                     10
                                                                 2                                                5                                                                                     2                                               2

                                                                 1                                                                                                                                      1                   /u/                                                            /w/
                                                                                           /u/                                  /w/
                                                                                                                  0                                                                                                                                     0
                                                                 0                                                                                                                                      0

                                                                 -1                                              -5                                                                                 -1                                                  -2
                                                                      0            5                     10           0               5                 10                                                  0           5                     10             0             5                     10
                                                                                                 Cepstral Coefficient                                                                                                                   Cepstral coefficient

                                                                                                                                                                                        qualitatively distinguished by the frequency component of
b. Formant Estimation of BODO and RABHA Phonemes                                                                                                                                        the vowel. Generally, three formants frequencies (F1, F2 and
                                                                                                                                                                                        F3) are considered for perception and discrimination of
Formant frequency                                                      is the distinguishing frequency                                                                                  vowels by a listener (Kewley, 1982, 1983). A variety of
components of human                                                   speech. It refers to specific resonance                                                                           approaches, such as formant tracking articulator model and
frequencies of vocal                                                  tract which have maximum energy                                                                                   auditory model have been used for the analysis and synthesis
concentration during                                                   the vowels utterance. It can be                                                                                  of speech. The formant tracking method, based on Linear
                                                                                                                                                                                                                                                                                                      5


@ 2012, IJCCN All Rights Reserved
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9


Predictive Code (LPC), has received considerable attention.
Based on digitalized technique, the entire frequency range is
divided into a fixed number of segment and each segment is                                       =                                                              (25)
represents a formant frequency. A 2nd order resonator for                                        The parameter             , determines the bandwidth of the
each segment k with a specific boundary is defined. A
predictor polynomial defined as the Fourier transform of the                                     resonator defined as negative (-) of           .               .The
corresponding 2nd order predictor is given by (Welling. I,                                       formant frequency is given by,
and Ney, II, 1998):
                                                                                                                                                              (26)
                                                                           (23)
Where        and      are real valued predictor coefficients.
Therefore, from equation (23) we get
                                                                           (24)

                                                         Table 3:    Formant Frequencies Estimation of BODO Words
                                                                              Formant frequency
          Vowe                   /o/                    /a/                   /i/                /e/                         /u/              /w/
          Female       F1       319.1                   380.3                 411.3              387.5                       249.6            292.7
                       F2       833.3                   1194.5                2409.8             2240.8                      997.7            1527.2
                       F3       3030.4                  3650.4                2911.8             3165.0                      3044.3           3165.3
          Male         F1       309.3                   343.8                 394.6              384.9                       244.7            206.4
                       F2       764.0                   1172.0                2341.6             2178.1                      837.5            1147.1
                       F3       2748.8                  2494.5                3002.4             3577.1                      3690.6           2486.9
          VC                    /or/(fire)                                    /ich/(pain)        /un/(back side)             /ul/(confuse)    /em/(bed)
                                                        /      /(I)
          Female       F1       326.4                   326.1                   293.3                 300.5                  347.2            311.4
                       F2       1623.4                  1717.5                  2371.3                1424.7                 2353.1           2452.7
                       F3       3023,8                  3006.2                  3455.9                3276.9                 2853.5           2765.3
          Male         F1       539.1                   714.0                   299.3                 280.2                  442.8            398.5
                       F2       2293.9                  2365.5                  2932.2                2240.0                 2544.9           1265.7
                       F3       3242.6                  3199.6                  3189.1                2636.7                 3350.7           2435.8
          CV                    /hw/(to give)           /bu/(to swell)          /ru/(to boil)                                /be/(this)       /gi/(to fear)
                                                                                                      /       /(to beat)
          Female       F1       320.7                   382.1                   311.1                 337.6                  354.6            334.8
                       F2       1687.9                  1661.1                  1623.5                1853.7                 1699.5           1617.9
                       F3       3120.24                 3077.1                  3445.5                2996.8                 3001.65          2947.7
          Male         F1       494.4                   690.1                   633.0                 375.5                  283.4            393.0
                       F2       2109.8                  2545.8                  2386.2                2536.2                 2250.1           2223.5
                       F3       3216.3                  3355.9                  3298.9                2842.9                 3220.0           3287.7
          CVC                   /san/(the sun)          /swb/(smoke)            /bar/(wind)           /lir/(to write)        /dwn/(to keep)   /thar/ (sure)
          Female       F1       285.5                   282.5                   298.5                 304.7                  352.6            276.1
                       F2       1800.6                  1966.4                  2657.89               2354.87                1471.0           2491.2
                       F3       3286.8                  3135.6                  3024.78               3254.67                3163.2           3155.5
          Male         F1       838.3                   727.5                   892.2                 745.3                  300.7            415.6
                       F2       1494.4                  1421.3                  1356.9                1293.2                 1238.3           1629.4
                       F3       3546.54                 3265.67                 3198.00               3354.52                3648.01          3674.98




                                                                                                                                                                       6


@ 2012, IJCCN All Rights Reserved
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9


                       Formant Frequencies estimation of 6 Bodo vowels for                                                                                        Formant Frequencies estimation of 6 Bodo vowels for male
                       female utterances                                                                                                                          utterances
                                                50                                                     50                                                                                    50                                                        40
                                                                                                                                /e/                                                                                             /o/                                                /e/
                                                                                                                                                                                                                                                       20
                                                                                     /o/
                                                  0                                                     0                                                                                        0
                                                                                                                                                                                                                                                        0

                                                -50                                                    -50                                                                                  -50                                                        -20
                                                      0         1000     2000       3000     4000            0   1000    2000     3000         4000                                                  0        1000      2000       3000      4000            0     1000    2000    3000    4000
                                                50                                                     20                                                                                    50                                                        40




                                                                                                                                                                             A p d (d )
                              A p d d )




                                                                                                                                                                              m litu e B
                               m litu e B
                                                                                     /a/                                        /u/                                                                                                 /a/                                            /u/
                                            (

                                                                                                        0                                                                                                                                              20
                                                  0                                                                                                                                              0
                                                                                                       -20                                                                                                                                              0

                                                -50                                                    -40                                                                                  -50                                                        -20
                                                      0         1000     2000       3000     4000            0   1000    2000     3000         4000                                                  0        1000      2000       3000      4000            0     1000    2000    3000    4000
                                                20                                                     50                                                                                    50                                                        40
                                                                                      /i/                                         /w/                                                                                                /i/                                            /w/
                                                  0                                                                                                                                                                                                    20
                                                                                                        0                                                                                        0
                                                -20                                                                                                                                                                                                     0

                                                -40                                                    -50                                                                                  -50                                                        -20
                                                      0         1000     2000       3000     4000       0        1000    2000     3000         4000                                                  0        1000      2000       3000      4000            0     1000    2000    3000    4000
                                                                                             Frequency(Hz)                                                                                                                                   Frequency(Hz)


                       F1-F2 plot shows the vowel triangle for male and female                                                                                    Formant frequency curves shows the distinction of formant
                       speaker of Bodo language.                                                                                                                  variation for V,VC,CV & CVC word structure.
                                                                                   F1-F2 for male & female vowel tringle                                                                                                           Change of fromant with v/vc/cv/cvc
                                        2600                                                                                                                                               40

                                        2400                                                                                    Red-Male                                                                                       V
                                                                             /i/
                                                                                                                                Blue-Female                                                30
                                        2200                                                                                                                                                                                    VC


                                        2000                                                                                                                                               20
                                                                                                                                                                                                                                             CV

                                        1800
                                                                                                                                                                                           10




                                                                                                                                                                  G in(d )
                                                                                                                                                                        B
                                                                                                                                                                                                                                                                          CV C
                       2 z)
                      F (H




                                        1600




                                                                                                                                                                   a
                                                                                                                                                                                            0
                                        1400

                                        1200                                                                                                                                               -10

                                        1000
                                                                                                                                               /a/                                         -20
                                            800           /u/

                                            600                                                                                                                                            -30
                                              200                      250             300            350          400              450               500                                        0             500        1000             1500    2000    2500             3000         3500     4000
                                                                                                    F1 (Hz)                                                                                                                                   Frequency (Hz)


F1-F2 plot shows the range formant frequencies of the CV,VC or CVC word structure of Bodo language mostly lies within the
range of the formant frequencies of the vowels.
                                                                                                                                                                                            Range of Formant Frequency
                                                                                                                                        2500



                                                                                                                                                                                                         VC
                                                                                                                                        2000


                                                                                                                                                                                                                                     CV
                                                                                                                                                                                                                                                  FV
                                                                                                                                        1500
                                                                                                                                                                                                                                    MV
                                                                                                                                 2 z)
                                                                                                                                F (H




                                                                                                                                        1000                                                                              CVC




                                                                                                                                         500

                                                                                                                                                                                                                                           MV-Male,vowel
                                                                                                                                                                                                                                           FV-Female,vowel
                                                                                                                                          0
                                                                                                                                          200               250                             300                 350            400            450            500
                                                                                                                                                                                                              F1 (Hz)




                                                                                                                              Table 4: Formant frequency
                                                                                                                                             Formant frequency
            Vowel                                               /o/                                              /a/                /i/                   /e/                                                                                                /u/                          /w/
            Female                              F1              640.3                                            283.5              280.4                 1040.8                                                                                             480.9                        340.5
                                                F2              2560.4                                           1480.3             2560.8                1384.4                                                                                             2360.1                       1080.2
                                                F3              3220.8                                           3600.2             2200.6                3151.8                                                                                             3211.4                       2720.4
            Male                                F1              620.2                                            243.8              301.3                 987.4                                                                                              504.5                        253.9
                                                F2              2154.7                                           1654.4             2251.8                2657.9                                                                                             2857.9                       2415.7
                                                F3              2876.1                                           2865,8             3985.8                3758.4                                                                                             3415.8                       2965.8
            VC                                                                                                                      /intcek/(this much) /ek/(to jump)                                                                                        /ut/(camal)
                                                                /ora    /(you are)                               /       /(I am)                                                                                                                                                          /r    /(length)
            Female                              F1              543.7                                            375.3              275.7                 765.2                                                                                              653.9                        392.8
                                                F2              1748.9                                           1682.9             2769.4                1765.6                                                                                             2015.9                       2438.3
                                                F3              3823.5                                           30165.5            3321.9                3546.9                                                                                             2976.9                       2657.9
            Male                                F1              643.3                                            987.5              276.9                 321.9                                                                                              431.9                        400.3
                                                F2              2396.9                                           2401.6             3001.8                2394.9                                                                                             2656,7                       1834.9
                                                F3              3242.6                                           3099.0             3548.2                2987.3                                                                                             3241.9                       2865.9
            CV                                                  /to/(hen)                                        /tsa/(to eat)      /mi/(vegetable)       /the/(fruit)                                                                                       //tcu/(thorn)
                                                                                                                                                                                                                                                                                          /a    /(shout)
            Female                              F1              465,9                                            3428                                354.9                                                              698.8                                387.03                       565.7
                                                F2              1874.5                                           2463.9                              1987.7                                                             1976.4                               1687.5                       176.5
                                                                                                                                                                                                                                                                                                            7


@ 2012, IJCCN All Rights Reserved
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9


                                          F3                   2976.4                                     2981.6                               3768,9                                      2885.6                        3415.6                           2986.3
            Male                          F1                   498.3                                      690.3                                541,7                                       367.2                         298.5                            391.2
                                          F2                   2183.5                                     2574,8                               3286.1                                      2653.0                        2261.3                           2695.3
                                          F3                   3216.3                                     3582.7                               3321.8                                      2976.2                        3139.7                           3271.6
            CVC                                                /tcok/(compound)                                                                /rin)(loan)                                                               /tbau/(owl)
                                                                                                          /na   (You                                                                       /ben                                                           /tsara
                                                                                                          are)                                                                             /(where)                                                       /(disease)
            Female                        F1                   276.4                                      265.8                                301.6                                       312.4                         299.7                            261.6
                                          F2                   1867.5                                     20001.8                              2782.5                                      2323.3                        1976.4                           2434.1
                                          F3                   3341.7                                     3875.4                               3054.6                                      3198.6                        2988.3                           3145.1
            Male                          F1                   845.2                                      698.4                                875.2                                       684.9                         301.5                            500.8
                                          F2                   1476.7                                     1501.0                               1401,9                                      1354.8                        1222.6                           1687.8
                                          F3                   3498.6                                     3315.8                               3176.0                                      3299.7                        3571.8                           3679.1

                  Formant Frequencies estimation of 6 Rabha Vowels for                                                                                  Formant Frequencies estimation of 6 Rabha vowels for male
                  female utterances                                                                                                                     utterances
                                          20                                                   20                                                                          20                                                40
                                                                                                                           /e/
                                                                             /o/                0                                                                              0                                             20
                                               0
                                                                                              -20                                                                          -20                                                0

                                        -20                                                   -40                                                                          -40                                              -20
                                                   0         1000     2000   3000    4000           0   1000     2000    3000        4000                                           0      1000   2000   3000     4000            0        1000   2000     3000     4000

                                          20                                                   20                                                                          20                                                40




                                                                                                                                                             A p d (d )
                                                                                                                                                              m litu e B
                               de )
                          plitu (dB




                                                                              /a/                                              /u/                                                                                           20
                                               0
                                                                                                0                                                                              0
                                        -20                                                                                                                                                                                   0
                        Am




                                                                                                                                                                           -20                                              -20
                                        -40                                                   -20                                                                                   0      1000   2000   3000     4000            0        1000   2000     3000     4000
                                                   0         1000     2000   3000    4000           0   1000     2000    3000        4000
                                                                                                                                                                           40                                                40
                                          20                                                   20
                                                                              /i/                                              /w/                                         20                                                20
                                               0                                                0                                                                              0                                              0

                                                                                                                                                                           -20                                              -20
                                        -20                                                   -20                                                                                   0      1000   2000   3000     4000            0        1000   2000     3000     4000
                                                   0         1000     2000   3000    4000           0   1000     2000    3000        4000
                                                                                                                                                                                                                  Frequency(Hz)
                                                                                      Frequency(Hz)



                  F1-F2 plot shows the vowel triangle for male and female                                                                               F1-F2 plot shows the range formant frequencies of the
                  speaker of Rabha language                                                                                                             CV,VC or CVC word structure of Rabha language mostly
                                                                             Vowel tringle for male & female speaker                                    lies within the range of the formant frequencies of the
                                               3000
                                                                                                                       Red-Male
                                                                                                                                                        vowels.
                                                                                                                       Blue-Female                                                                          Range of Formant frequency
                                                               /i/                                                                                                                  3000
                                               2500                                                                                                                                                                                          MV-Male,vowel
                                                                                                                                                                                    2800                                                     FV-Female,Vowel

                                                                                                                                                                                    2600

                                               2000                                                                                                                                 2400
                                                                                                                                                                                                                                      CV
                                      F2(Hz)




                                                                                                                                                                                    2200
                                                                                                                                                                                                                                                     MV
                                                                                                                                                                             (Hz)




                                                                                                                                                                                    2000                        CVC
                                               1500
                                                                                                                                                                           F2




                                                                                                                         /a/                                                                                                                         FV
                                                                                                                                                                                    1800
                                                                                                                                                                                                                                      VC
                                                                                                                                                                                    1600
                                               1000
                                                                /u/
                                                                                                                                                                                    1400

                                                                                                                                                                                    1200

                                                   500                                                                                                                              1000
                                                         0                          500                   1000                          1500                                               0                500                            1000                   1500
                                                                                            F1 (Hz)                                                                                                                      F1(Hz)


                                                                                                                                                             minimum with respect to vowel /o/. In case of female
3.RESULTS AND DISCUSSION                                                                                                                                     speaker, the maximum variation of cepstral co-efficient is
                                                                                                                                                             found in case of vowel /o/ and minimum with respect to
Depending on the analysis on cepstral features and formant                                                                                                   vowel /e/.
frequencies of Bodo and Rabha phonemes and words the
following observations were made-Significant variation of                                                                                                    Significantly, cepstral coefficients of Bodo vowels for frame
cepstral coefficients are observed among the Bodo vowels as                                                                                                  nos: 2, 3, 6 & 8 have shown distinctive characteristic
shown in Table-1. The cepstral variation is found to be                                                                                                      (Figure-4) for male and female speaker. The variation of the
maximum with respect to vowel /o/ and minimum                                                                                                                cepstral coefficients for male is very irregular in contrast to
corresponding to vowel /u/, in case of male speakers.                                                                                                        the stable variation of female cepstral coefficients. The same
Similarly, for female Bodo speakers, the maximum variation                                                                                                   phenomenon is also observed in case of Rabha vowels also,
of cepstral measure is found corresponding to vowels /o/ and                                                                                                 but in this case the frame numbers are different i.e. frame no:
minimum in case of /i/.                                                                                                                                      9, 14, 16 and 17 (Figure-7). This observation may be helpful
                                                                                                                                                             in sex determination for both Bodo and Rabha speakers.
In case of Rabha vowels, i.e., /o/, /a/, /i/, ./e/,, /u/ and /w/ for                                                                                         The range of variation of cepstral coefficients for Bodo and
both male and female speakers the range of variation of the                                                                                                  Rabha male is found within the range of 3.8177
cepstral coefficient (Table-2) is found to be maximum in                                                                                                     >CBodo>1.1523 and 8.1329>CRabha>2.0579 respectively.
case of male speakers with respect to vowel /u/ and                                                                                                          The range of variation for female is found
                                                                                                                                                                                                                                                                           8


@ 2012, IJCCN All Rights Reserved
Jyotismita Talukdar et al., International Journal of Computing, Communications and Networking, 1(1), July – August, 1-9


1.9578>CBodo>0.9276 and 7.6546>CRabha>2.4127. i.e. the                                           3.    Borz. Porat. A course in digital Signal Processing,
variation of cepstral features for Bodo vowels is less (Male-                                          John Willy & Sons. 1997.
2.6654; Female-1.0302)with respect to the Rabha
vowels(Male-6.0750;Female-5.2419) i.e., the former is                                            4.    Proakis, J.G. and Manolakis, D.G. Digital Signal
stable as compared to the latter.                                                                      Processing Principles, Algorithm and Applications,
                                                                                                       Pearson edition, Third Indian reprint 2004.
The Figure 10 and Figure 15 represent the extremes of
formant locations in the F1-F2 plane for both Bodo and                                           5.    Kewley-Port, D. Measurement of formant transitions
                                                                                                       in naturally produced stop consonant-vowel
Rabha vowels. It is found that the formant locations for /u/                                           syllables, Journal of the Acoustical Society of America,
(low F1, low F2), /i/ (low F1, high F2) and /a/(high F1, low                                           72, pp. 379-389, 1982.
(low F1, low F2), /i/ (low F1, high F2) and /a/(high F1, low
F2) with other vowels are placed with respect to the triangle                                    6.    Kewley-Port, D. Time-varying features as correlates
vertices.                                                                                              of place of articulation of stop consonants, Journal of
                                                                                                       the Acoustical Society of America, 73, pp. 322-335,
The Figure 12 and Figure 16 have shown that the formant                                                1983.
frequencies of the selected word sets for both Bodo and
Rabha lies within the range of the formant frequencies of the                                    7.    Willing I., and Ney, II. Formant Estimation for
isolated vowels. The investigation have shown that (Table-3                                            Speech Recognition, IEEE Transactions on Speech and
& 4) the range of formant frequency is maximum in case of                                              Audio Processing, Vol 6. pp.-36-48,1998.
isolated vowels, but when the vowels are placed in the
nucleus of a structure like CV, VC or CVC, the formant                                           8.    Talukdar P.H; 2010. Speech production, Analysis and
frequency decreases.                                                                                   Coding, Lambert Publication, Germany 2010.

ACKNOWLEDGEMENT

We highly acknowledge the Ministry of Communication &
Information Technology (MIT), New Delhi, Govt. of India,
for providing us the relevant information while preparing the
manuscript of this paper.

REFERENCES

1.    Rabiner, L.R and B. H. Juang. Fundamentals of
      Speech Recognition, Prentice-Hall, Englewood Cliff,
      New Jersy, 1993.

2.    A.M. Noll. Spectrum Pitch Determination, J. Acoustic
      Society. A.M. Vol.41. pp.293-309, Feb.1967




                                                                                                                                                              9


@ 2012, IJCCN All Rights Reserved

								
To top