Text to Phoneme Transcription Capable to Detect Ezafe and Homographs for Persian Speech Synthesis

Please enable javascript in your browser.

Oskouipour, Navid | 2011

570 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 43030 (19)
University: Sharif University of Technology
Department: Computer Engineering
Advisor(s): Sameti, Hosein
Abstract:
In Persian language, there are some special uncertainties and complexities regarding the extraction of phonemes due to its flexible structure. In this work, we try to solve two problems caused by this issue, homographs disambiguation and Ezafe recognition in Persian text to speech systems.
In this work, two methods are presented for solving the two problems. The first one is a pattern recognition method based on Aho-Corasick algorithm that has the best performance according to the execution time. The second method is a hidden Markov model tagger based on Viterbi algorithm that uses bigram and trigram statistics. These methods are run on 10 million word corpus and the results are compared to the results of previous works on this area.
The trigram HMM method has the best performance on big corpus for Ezafe recognition (accurancy over 91%). This method also has the best result on solving the problem of homographs disambiguation. Furthermore, the weakness of previous definition of Ezafe construction is shown with some examples and a new definition is presented
Keywords:
Text-to-Speech Transform ; Homographs ; Ezafe ; Ezafe Construction

No TOC