Malay text-to-speech

• المؤلف الرئيسي: Mohammed Jan Solihn Donyada.
• مؤلفون آخرون: Soon Ing Yann
• التنسيق: Final Year Project
• اللغة: English
• منشور في: 2011
• الموضوعات: DRNTU::Engineering::Electrical and electronic engineering
• الوصول للمادة أونلاين: http://hdl.handle.net/10356/46425
• المؤسسة: Nanyang Technological University

Speech synthesis has proven to be an important assistive technology tool and its application in this aspect is very widespread. Text-to-speech has commonly been used as screen readers for people with visual impairment. Text-to-speech systems today are also used by people with reading difficulties, pre-literate children as well as people with severe speech impediments. This report details the development of a basic Text-to-Speech program for the Malay Language using MATLAB and GUIDE, with a database of recorded words and affixes using Audacity. A CTTS (Concatenative Text-to-Speech) approach was taken, with an extensive database of prerecorded natural voice segments made up of words and affixes. Altered root words that change form with the addition of certain prefixes have been included within the database, in addition to the original root word. A program that was able to dissect a Malay word into its component affixes and root was created. A morphological analyzer was hard-coded to perform the scanning and extraction of the affixes within a word. Natural sounding speech synthesis was achieved.