Third International Conference on Spoken Language Processing (ICSLP 94)
A practical text-to-speech system developed for Hindi language is discussed. The linguistic component includes a rule base and a segment base. The rule base converts a text input into an appropriate phoneme output of the language. For this purpose, various layers of rules are incorporated. The segment base consists of a library of digitized speech segments, of various lengths, demarcated on certain principles. Other part of the software includes modules for communication on serial link, database management and speech replay through the D/A converter. The hardware part of the system includes CPU, memory and the necessary I/O interfaces. Applications for the use of visually handicapped and illiterates are discussed.
Bibliographic reference. Bhaskararao, P. / Peri, Venkata N. / Udpikar, Dr. Vishwas (1994): "A text-to-speech system for application by visually handicapped and illiterate", In ICSLP-1994, 1239-1242.