Determination of input parameters of the neural network model, intended for phoneme recognition of a voice signal in the systems of distance learning
Abstract
The article is devoted to the problem of voice
signals recognition means introduction in the system of distance
learning. The results of the conducted research determine the
prospects of neural network means of phoneme recognition.
It is also shown that the main diculties of creation of the
neural network model, intended for recognition of phonemes
in the system of distance learning, are connected with the
uncertain duration of a phoneme-like element. Due to this
reason for recognition of phonemes, it is impossible to use
the most eective type of neural network model on the basis
of a multilayered perceptron, at which the number of input
parameters is a xed value. To mitigate this shortcoming, the
procedure, allowing to transform the non-stationary digitized
voice signal to the xed quantity of mel-cepstral coecients,
which are the basis for calculation of input parameters of
the neural network model, is developed. In contrast to the
known ones, the possibility of linear scaling of phoneme-
like elements is available in the procedure. The number of
computer experiments conrmed expediency of the fact that
the use of the oered coding procedure of input parameters
provides the acceptable accuracy of neural network recognition
of phonemes under near-natural conditions of the distance
learning system. Moreover, the prospects of further research in
the eld of development of neural network means of phoneme
recognition of a voice signal in the system of distance learning
is connected with an increase in admissible noise level. Besides,
the adaptation of the oered procedure to various natural
languages, as well as to other applied tasks, for instance, a
problem of biometric authentication in the banking sector, is
also of great interest.
Full Text:
PDFReferences
V. Mikhaylenko Neural network models and methods of recognition
of phonemes in a voice signal in the system of distance learning:
[Monograph] / V. M. Mikhailenko, L. O. Tereykovskaya, I.
A. Tereykovsky., B. B. Akhmetov. - K .: CP "Komprint", 2017.-
p.
A Najib, A Basari, A Pee, M Daimon, A Rahman, L Tahir
ONLINE PERFORMANCE DIALOGUE SYSTEM MODEL
(e-DP): A REQUIREMENT ANALYSIS STUDY AT BATU
PAHAT DISTRICT EDUCATION OFFICE Journal of Theoretical
and Applied Information Technology. 31st December
Vol.95. No 24 P. 6699 6706.
A. Mosa, M. Mahrin, R. Yuso A SYSTEMATIC LITERATURE
REVIEW OF TECHNOLOGICAL FACTORS FOR ELEARNING
READINESS IN HIGHER EDUCATION. Journal
of Theoretical and Applied Information Technology. 30th
November 2016. Vol.93. No.2. P. 500 521.
I. Veritawati, I. Wasito, T. Basaruddin TEXT INTERPRETATION
USING A MODIFIED PROCESS OF THE ONTOLOGY
AND SPARSE CLUSTERING. Journal of Theoretical
and Applied Information Technology 15th March 2017. Vol.95.
No 5. P. 1019-1028.
A.Kadir, A. Yauri AUTOMATED SEMANTIC QUERY FORMULATION
USING MACHINE LEARNING APPROACH.
Journal of Theoretical and Applied Information Technology.
th June 2017. Vol.95. No 12. P. 2761-2775.
J. Park, J. Yoon, Y. Seo, G. Jang SPECTRAL ENERGY
BASED VOICE ACTIVITY DETECTION FOR REAL-TIME
VOICE INTERFACE. Journal of Theoretical and Applied
Information Technology. 15th September 2017. Vol. 95 No17.
P. 4304-4312.
A Agranovsky, D. Lednov Theoretical aspects of algorithms
for processing and classifying speech signals. - M .: Radio and
Communication, 2004. - Ch. 1. 164 c.
L. Babenko, D. Subbotin, V. Fedorov, P. Yurkov DEFINITION
OF THE BORDERS BETWEEN THE FONEMAS BY A
NEUROET NETWORK METHOD. Izvestiya Southern Federal
University. Technical science. 2003 4 Òîì33. Pp. 321-323.
T. Kartbayev, B. Akhmetov, A. Doszhanova, K. Mukapil,
A. Kalizhanova, G. Nabiyeva, L. Balgabayeva, F. Malikova
DEVELOPMENT OF A COMPUTER SYSTEM FOR IDENTITY
AUTHENTICATION USING ARTIFICIAL NEURAL
NETWORKS. Image Analysis & Stereology, 10.5566/ias.1612.
V.36, 1, 2017.
O. Fedyaev, I. Bondarenko Neural network algorithm for
speaker-independent recognition of speech phonemes. USIM,
, No. 4 C. 41- 50.
B. Meyer , T. Jurgens, T. Wesker, T. Brand, B. Kollmeier
Human phoneme recognition depending on speech-intrinsic
variability. J Acoust Soc Am. 2010 Nov;128(5):3126-41.
Y. Qian, M. Bi, T. Tan, K. Yu, "Very deep convolutional neural
networks for noise robust speech recognition," in IEEE/ACM
Trans. Audio Speech Language Process. , vol. 24, no. 12, pp.
-2276, 2016.
V. Lila, E. Puchkov Methodology of training a recurrent arti-
cial neural network with dynamic stack memory. International
magazine "Software products and systems", Tver, 4, 2014 p.
[on pages 132-135].
Understanding LSTM Networks Posted on August 27,
(http://colah.github.io/posts/2015-08-Understanding-
LSTMs/) .
A. Waibel, T. Hanazawa, G. Hinton, K. Shikano, K. Lang
¾Phoneme Recognition Using Time - Delay Neural Networks¿,
IEEE Transactions On Accoustics, Speech And Signal Processing,
Vol. 37, 1989.
M. Gusev Methods and models of recognition of Russian speech
in information systems: dis. ... doctors of techn. Sciences:
13.01 / MN Gusev - St. Petersburg, 2014. - 378 p.
I. Tereykovskii Optimization of the structure of a two-chirped
perceptron, possible distribution of fertility of anomalous inuences
of experimental parameters of computer technology / IA
Tereykovskii // Scientic and technical journal "Management
of branching of folded systems" Kiev. National University of
Architecture. - 2011. - Vol. 5. - S. 128-131.
I. Boykov, A. Ivanov, D. Kalashnikov ALGORITHM OF
THE CONSTRUCTION OF THE STATISTIC DISCRETECONTINUOUS
DESCRIPTION OF THE DURATION OF
THE AUDIO SOURCES OF THE INCREASED SPEECH OF
THE DICTOR. News of higher educational institutions. The
Volga region. 4 (36), 2015 p.64-76
Refbacks
- There are currently no refbacks.
International Journal of Electronics and Telecommunications
is a periodical of Electronics and Telecommunications Committee
of Polish Academy of Sciences
eISSN: 2300-1933