Grapheme to Phoneme (G2P) conversion for Bangla

Basu, Joyanta; Basu, Tulika; Mitra, Mridusmita; Mandal, Shyamal Kr. Das

doi:10.1109/icsda.2009.5278373

Cited by 21 publications

(7 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…BGLM was trained on 1,50,000 unique Bangla words collected from 4,80,000 Bangla sentences of news domain where digits are not considered as unique words. EPLM and BPLM were trained on the phoneme sequences generated using Bangla G2P [7] module after applying on same training data of EGLM and BGLM respectively. The proposed methodology had been evaluated on sentences containing 23,086 unique words collected from blog and social networking websites written by metropolis educated Bangla students.…”

Section: Resultsmentioning

confidence: 99%

“…Moreover, the complexity may increases if such an English word is used with Bangla suffixes. To overcome these challenges, our English word detection methodology utilizes Bangla Morphological Analyzer (BMA) and Bangla Grapheme to Phoneme Convertor (BGPC) [7]. Initially the input sentence is tokenized in words and then each word is passed through BMA to find out the root word.…”

Section: The Structure Of Bangla Grammar Does Not Change Inmentioning

confidence: 99%

“…If frequency of the word is very low or not found in the dictionary, it is assumed that the word may be an English word or Bangla OOV word. Now, phoneme of the words are extracted using Grapheme to Phoneme (G2P) convertor [7] and joint probability of the generated phoneme sequences and word are estimated using two different Phoneme Language Model (PLM) and Grapheme Language Model (GLM). PLM estimates joint probability of phoneme sequences by considering consecutive trigram frequencies.…”

Section: The Structure Of Bangla Grammar Does Not Change Inmentioning

confidence: 99%

See 2 more Smart Citations

Automatic detection of English words in Benglish text: A statistical approach

Kundu

Chandra

2012

2012 4th International Conference on Intelligent Human Computer Interaction (IHCI)

View full text Add to dashboard Cite

Code-mixing and code-switching create challenges in the field of natural language processing applications like Machine Translation and Speech-to-Speech Translation. Detection of foreign words is very much essential for smooth processing of natural language.A statistical language independent approach for automatic detection of foreign words in mixed language has been introduced in this paper. Initially, the proposed approach has been applied on Benglish text which is combination of Bangla text contains English words. The methodology can be easily adopted for other languages where such code mixing exists. The proposed approach yields an accuracy of 71.82% when tested on sentences collected from Bangla blogs and social networking websites.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: The Structure Of Bangla Grammar Does Not Change Inmentioning

confidence: 99%

Section: The Structure Of Bangla Grammar Does Not Change Inmentioning

confidence: 99%

See 1 more Smart Citation

Automatic detection of English words in Benglish text: A statistical approach

Kundu

Chandra

2012

2012 4th International Conference on Intelligent Human Computer Interaction (IHCI)

View full text Add to dashboard Cite

show abstract

“…That all the seven Bangla oral vowels have their phonemically contrastive nasal counterparts has been widely reported in earlier literature (Morshed, 1972;Alam et al, 2008;Barman, 2011;Thompson, 2012;Basu, Basu, Mitra, & Mandal, 2009;Shamim, 2011;Huq, 2002;Sarkar, 2004;Khan, 2008). As Morshed (1972) described, "… the number of nasalized vowels [in SCB] is equal to that of their oral counterparts, though the frequency of occurrence of nasalized vowels is far less than that of oral vowels" (p. 62).…”

Section: Nasal Vowels In Banglamentioning

confidence: 94%

Phonemic status of Bangla nasal vowels: A corpus study

Islam¹

2018

ALA

View full text Add to dashboard Cite

This study investigated the phonemic status of the nasal vowels in Bangla (aka Bengali). It has been claimed for decades that all the seven monophthongal oral vowels in Bangla have phonemically contrastive nasal counterparts; however, an in-depth investigation of the status of nasality for all the vowels is lacking in the current literature. With a phoneme dictionary build from a text corpus of 8 (eight) million word-tokens and about 275 thousand word-types, this study investigated whether all the oral vowels have phonemically contrastive nasal vowels. Findings revealed that only five of the seven monophthongal vowels form phonemically contrastive relationships with their nasal counterparts; nasality in /ae/ and /ɔ/ are not contrastive phonemically. Keywords: nasal vowels; phonemic status; text corpus; Bangla/Bengali PovzetekČlanek preučuje fonemski status nosnih samoglasnikov v bengalščini (dialekt aka). Skozi desetletja je prevladovalo mnenje, da ima v bengalščini vseh sedem enostavnih samoglasnikov fonemsko kontrastivni nosni par, vendar pa to mnenje ni bilo podprto s poglobljenimi raziskavami. S pomočjo fonemskega slovarja, ki je bil zgrajen iz besedilnega korpusa z 8 milijoni pojavnic -od tega je približno 275 tisoč različnic -smo preverili, ali imajo vsi osnovni enostavni samoglasniki tudi svoje fonemsko kontrastivne nosne pare. Rezultati nakazujejo, da je takšnih parov v kontrastivni distribuciji samo 5 in da nosna /ae/ and /ɔ/ nista fonemsko kontrastivna.

show abstract

“…But in case of Bengali language, LTS rules are not as simple as other Indian languages. In Basu et al (2009), Ghosh et al (2010, simple LTS rules for Bengali have been derived and incorporated in speech synthesis and speech recognition systems.…”

Section: Deriving Letter To Sound Rulesmentioning

confidence: 98%

Development of syllable-based text to speech synthesis system in Bengali

Narendra

Rao

Ghosh

et al. 2011

Int J Speech Technol

View full text Add to dashboard Cite

This paper presents the design and development of unrestricted text to speech synthesis (TTS) system in Bengali language. Unrestricted TTS system is capable to synthesize good quality of speech in different domains. In this work, syllables are used as basic units for synthesis. Festival framework has been used for building the TTS system. Speech collected from a female artist is used as speech corpus. Initially five speakers' speech is collected and a prototype TTS is built from each of the five speakers. Best speaker among the five is selected through subjective and objective evaluation of natural and synthesized waveforms. Then development of unrestricted TTS is carried out by addressing the issues involved at each stage to produce good quality synthesizer. Evaluation is carried out in four stages by conducting objective and subjective listening tests on synthesized speech. At the first stage, TTS system is built with basic festival framework. In the following stages, additional features are incorporated into the system and quality of synthesis is evaluated. The subjective and objective measures indicate that the proposed features and methods have im-

show abstract

Grapheme to Phoneme (G2P) conversion for Bangla

Cited by 21 publications

References 5 publications

Automatic detection of English words in Benglish text: A statistical approach

Automatic detection of English words in Benglish text: A statistical approach

Phonemic status of Bangla nasal vowels: A corpus study

Development of syllable-based text to speech synthesis system in Bengali

Contact Info

Product

Resources

About