Implementation of Text To Speech for Marathi Language Using Transcriptions Concept

Sangramsing N. Kayte et al. Int. Journal of Engineering Research and Applications www.ijera.com
ISSN: 2248-9622, Vol. 5, Issue 11, (Part - 5) November 2015, pp.33-36
www.ijera.com 33 | P a g e
Implementation of Text To Speech for Marathi Language Using
Transcriptions Concept
Sangramsing N. Kayte1
, Monica Mundada1
, Dr. Charansing N. Kayte2
,
Dr.Bharti Gawali*
1,3
Department of Computer Science and Information Technology Dr. Babasaheb Ambedkar Marathwada
University, Aurangabad
2
Department of Digital and Cyber Forensic, Aurangabad, Maharashtra
ABSTRACT
This research paper presents the approach towards converting text to speech using new methodology. The text to
speech conversion system enables user to enter text in Marathi and as output it gets sound. The paper presents
the steps followed for converting text to speech for Marathi language and the algorithm used for it. The focus of
this paper is based on the tokenisation process and the orthographic representation of the text that shows the
mapping of letter to sound using the description of language’s phonetics. Here the main focus is on the text to
IPA transcription concept. It is in fact, a system that translates text to IPA transcription which is the primary
stage for text to speech conversion. The whole procedure for converting text to speech involves a great deal of
time as it’s not an easy task and requires efforts.
Keywords - IPA, Orthographic representation, phonetic, tokenization, transcription.
I. INTRODUCTION
The In today’s life everyone wants that the
present computer systems should behave like humans
and proved to be user friendly. Even many of the
great research scholars had dreamed of involving the
machines in every facet of human life. With the
growth of the power of computing machines their
applications in modern day to day life is also rising
[1]. This paper depicts one of the modern
technologies that is the technology of text to speech
convertor [2]. This paper presents the frame work of
text to speech conversion methodology. The text to
speech system has many applications that are useful
in our life such as for video conferencing, helping
people with reading difficulty, for visual aids process,
Elementary and Applied Research,
Telecommunications and Multimedia, Educational
State, Screen reader programs for the visually
impaired [3][4].
In this paper we have presented the linguistic
features of Marathi text stating its phonetic
representation. Through the knowledge of language’s
phonological orthographic we can develop a more
promising text to speech conversion [4][5]. The first
step in our text to speech conversion task is having
the text tokenisation and to determine the token to
sound concept.
II. OVERVIEW OF TEXT TO SPEECH
SYSTEM
The text to speech conversion is carried out with
the processing of text normalization (text
preprocessing), text analysis, syllabification, prosody
generation and finally the speech production
[6][7][15]. The process of text analysis, text
normalization and prosody generation all come under
natural language processing while the speech
generation is a part of digital processing system. A
general view of the system is depicted in the Fig.1
Fig 1: Overview of TTS [8]
The text to speech conversion involves a number
of steps that are to be followed. The first step
includes Tokenisation/Text preprocessing where the
inputted text is broken down into smallest unit i.e.
tokens. Tokenisation also includes the expansion of
abbreviations.
The second step involves Syllabification process
includes the combination of the small units to
RESEARCH ARTICLE OPEN ACCESS

represent them as phonemes that identify the sound
associated with each unit and we also called them as
syllables.
In Marathi seven types of syllables are
recognized [9][10].These syllable types are: V, VC,
CV, VCC, CVC, CCVC and CVCC; where V and C
represent vowel and consonant respectively. Out of
these seven types, occurrence of last two syllable
types having sound clusters, is very rare in Marathi.
The third step is Speech Synthesizer which involves
the generation of speech as output. There are various
techniques that can be used for synthesizing speech
[10]. The speech can be produced by concatenating
different pieces of recorded speech from the
database. Also, speech can be produced from
concatenation of phonemes, diphones, and unit
selection techniques and by Formant synthesis where
the synthesized speech output is created using
additive synthesis and an acoustic model (physical
modelling synthesis).Parameters such as fundamental
frequency, voicing, and noise levels are varied over
time to create a waveform of artificial speech.
III. ARCHITECTURE AND
IMPLEMENTATION
Fig-2: Architecture of TTS
The entire process of converting text to speech
includes analysis and synthesis [11] [16-19].
Although analysis part is similar for every techniques
used for text to speech conversion, the difference is
only in the synthesis part. The analysis part deals
with determining the orthographic description of a
particular language.
The implementation of text to speech conversion
goes through various steps as depicted in Fig-2. The
first step identifies the text for the processing. The
text normalization is carried out. After the text has
been normalized, the syllabification process is carried
out where the input text is broken down into smallest
unit called syllables
A. Text to IPA transcription
The phonetization is done which is the process of
representing vocal sounds by signs and written
characters. There are number of phonetic sounds that
represent a letter of particular language. These
phonetic sounds along with their written
representation can be combined to generate sound
using various speech synthesis tools.
The orthographic representation of text enables to
represent the syllables with phonetic representation
[12]. The orthographic set for a language consists of
letter to sound rules. These rules define the function
mapping of sequence of letter to sound segments.
Specifically, the orthographic representation
determined for a language, specifies the initial stages
for text to speech conversion [13].The phonetic
notation can be represented for Marathi language
using an IPA chart (International Phonetic
Alphabet).The IPA chart defines is an alphabetic
system of phonetic notation based on some Latin
alphabets [14]. They can be used for representation
of sounds of spoken language[16-19].
The principal vowels for Marathi can be depicted
from Fig-3. The principal vowels are symmetrically
distributed on a standard vowel chart: three front
vowels, two central vowels, and three back vowels.
The three back vowels are rounded [14].
Fig-3: Vowel chart
B) Consonants IPA chart
Representations of consonant sounds outside of
the core set are created by adding diacritics to letters
with similar sound values.Fig-4 represent the
consonants and their IPA symbols[1][3] [9].
Fig-4: IPA chart for consonants

The methodology followed here translates the text to
IPA transcription using some defined rules. These
rules are used for determining the letter orthographic
features. These orthographic features can further be
used by the speech synthesis tool to generate
sound[4][16-19].
IV. CONCLUSION
In this paper, we have described how the
phonological attributes of Marathi orthography may
be used to develop a text-to-speech system. The
phonetic nature of the language was surveyed, and
the set of rules for deriving these sounds using the
orthography were made explicit and carefully
deconstructed. Although the work done is not
complete as it has been done until determining the
linguistic features of Marathi text while the final
signal processing aspect of the task are un explored.
In the near future we would love to complete the text-
to-speech system, the text to speech system that is
able to be useful to the people in need.
REFERENCES
[1] Sangramsing Kayte, Monica Mundada
"Study of Marathi Phones for Synthesis of
Marathi Speech from Text" International
Journal of Emerging Research in
Management &Technology ISSN: 2278-
9359 (Volume-4, Issue-10) October 2015
[2] Sangramsing N.kayte “Marathi Isolated-
Word Automatic Speech Recognition
System based on Vector Quantization (VQ)
approach” 101th Indian Science Congress
Jammu University 03th Feb to 07 Feb 2014
[3] Sangramsing Kayte, Monica Mundada, Dr.
Charansing Kayte "Di-phone-Based
Concatenative Speech Synthesis System for
Hindi" International Journal of Advanced
Research in Computer Science and Software
Engineering -Volume 5, Issue 10, October-
2015
Charansing Kayte “Di-phone-Based
Concatenative Speech Synthesis Systems for
Marathi Language” OSR Journal of VLSI
and Signal Processing (IOSR-JVSP)
Volume 5, Issue 5, Ver. I (Sep –Oct. 2015),
PP 76-81e-ISSN: 2319 –4200, p-ISSN No. :
2319 –4197 www.iosrjournals.org
[5] Monica Mundada, Sangramsing Kayte, Dr.
Bharti Gawali "Classification of Fluent and
Dysfluent Speech Using KNN Classifier"
International Journal of Advanced Research
in Computer Science and Software
Engineering Volume 4, Issue 9, September
2014
[6] Simon King, “An introduction to statistical
parametric speech synthesis”, Indian
Academy of Sciences, October 2011, pp.
837–852.
Charansing Kayte "Performance Calculation
of Speech Synthesis Methods for Hindi
language IOSR Journal of VLSI and Signal
Processing (IOSR-JVSP) Volume 5, Issue 6,
Ver. I (Nov -Dec. 2015), PP 13-19e-ISSN:
2319 –4200, p-ISSN No. : 2319 –4197
[8] Sangramsing Kayte, Dr. Bharti Gawali
“Marathi Speech Synthesis: A review”
International Journal on Recent and
Innovation Trends in Computing and
Communication ISSN: 2321-8169 Volume:
3 Issue: 6 3708 – 3711
Charansing Kayte "A Corpus-Based
Concatenative Speech Synthesis System for
Marathi" IOSR Journal of VLSI and Signal
2319 –4200, p-ISSN No. : 2319 –4197
Charansing Kayte "A Marathi Hidden-
Markov Model Based Speech Synthesis
System" IOSR Journal of VLSI and Signal
2319 –4200, p-ISSN No. : 2319 –4197
[11] Dhvani-TTS System for Indian
Languages,(http://dhvani.sourceforge.net),
2001.
[12] Prof. Ian Maddieson,”Applied Phonetics:
Portuguese Text-to-Speech”, University of
California, Berkeley Linguistics 110: May
16, 2003.
Charansing Kayte "Implementation of
Marathi Language Speech Databases for
Large Dictionary" IOSR Journal of VLSI
and Signal Processing (IOSR-JVSP)
Volume 5, Issue 6, Ver. I (Nov -Dec. 2015),
PP 40-45e-ISSN: 2319 –4200, p-ISSN No. :
2319 –4197
[14] International Phonetic Association (IPA),
Handbook
[15] Sangramsing Kayte, Monica Mundada,
Santosh Gaikwad, Bharti Gawali
"PERFORMANCE EVALUATION OF
SPEECH SYNTHESIS TECHNIQUES
FOR ENGLISH LANGUAGE "
International Congress on Information and
Communication Technology 9-10 October,
2015
[16] Sangramsing Kayte, Monica Mundada,Dr.
Charansing Kayte” Speech Synthesis
System for Marathi Accent using
FESTVOX” International Journal of

Computer Applications (0975 – 8887)
Volume 130 – No.6, November2015
Charansing Kayte “Screen Readers for
Linux and Windows – Concatenation
Methods and Unit Selection based Marathi
Text to Speech System” International
Journal of Computer Applications (0975 –
8887) Volume 130 – No.14, November 2015
Charansing Kayte “ Performance Evaluation
of Speech Synthesis Techniques for Marathi
Language “ International Journal of
Computer Applications (0975 – 8887)
Volume 130 – No.3, November 2015
[19] Sangramsing Kayte, Monica Mundada,
Jayesh Gujrathi, “ Hidden Markov Model
based Speech Synthesis: A Review”
International Journal of Computer
Applications (0975 – 8887) Volume 130 –
No.3, November 2015

Implementation of Text To Speech for Marathi Language Using Transcriptions Concept

More Related Content

What's hot (15)

Viewers also liked (20)

Similar to Implementation of Text To Speech for Marathi Language Using Transcriptions Concept (20)

Recently uploaded (20)

Implementation of Text To Speech for Marathi Language Using Transcriptions Concept