New Text-to-Speech System Enhances Yoruba Language Accessibility Online

Researchers have developed TTSYoruba, a rule-based concatenative diphone speech synthesizer specifically for the Yoruba language. This system is integrated into YorubaName.com, an online dictionary of Yoruba names, demonstrating a practical application for preserving and promoting African languages in the digital sphere.

The synthesizer converts tone-marked Yoruba text into audio using a meticulously crafted phonological rule system and a database of 651 diphone units. This addresses the complex tonal nature of Yoruba, including the disambiguation of nasal sounds and the generation of contextual rising and falling tones, which are crucial for accurate pronunciation and meaning.

Furthermore, the project introduces an orthographic innovation by standardizing the use of the caron and circumflex as single-vowel contour tone markers. These symbols have been incorporated into the system's text normalization process and the WriteYoruba keyboard tool, simplifying the input of accurate Yoruba tones for users.

The development of TTSYoruba represents a significant advancement for low-resource African languages in the field of Natural Language Processing. By providing a high-quality speech synthesis tool, it enhances accessibility and usability for Yoruba speakers and learners, contributing to the digital presence and cultural preservation of the language.