Transfer and Adaptation of Text-to-Speech Technology to the Bulgarian Language and its Application in Cultural Tools with Emphasis on Accessibility

Start Date: 01/10/2005
End Date: 30/09/2007
Project Leader: Raptis Spyros

The aim of the project is to transfer text-to-speech (TtS) technology and to adapt it to the Bulgarian language for developing a speech synthesis engine for Bulgarian. An important by-product will be the collection and processing of Bulgarian corpora (text and spoken corpora). An additional goal of the project will be to exploit the TtS engine for speech-enhancing a cultural web portal that will be developed, making it more accessible to visually impaired people.

State-of-the-art technologies and a set of preexisting systems and tools will be employed for the development of the TtS system and adapted to the specific characteristics of the Bulgarian language. It should be noted that up to the starting time of the project there was no TtS system widely available for the Bulgarian language. The development of the web portal will be based on technology already developed in the context of a previously R&D project.

Text-to-speech is widely considered as a key enabling technology worldwide. Language technology tools are now an indispensable part in various research and development areas but also in educational technology. The potential application areas are numerous and span from automated personal aids, to automated announcement systems, call centers, telephone services and so on. However, one of the most important application areas is the development of assistive tools and services for disabled people.

The Institute for Language and Speech Processing / R.C. “Athena” (ILSP / R.C. “Athena”) that is managing the project, has a widely recognized experience in the design of language technology solutions. ILSP / R.C. “Athena” has been a pioneer in text-to-speech for Greek language, having developed 3 generations of TtS engines up to date, based on different methodologies. ILSP / R.C. “Athena”’s current TtS engine offers a synthetic voice of very high intelligibility and quality.

The activities of the project will revolve around three main axes:

  • the collection and processing of language resources for the Bulgarian language, such as text and spoken corpora,
  • the use and adaptation of existing language technologies and tools for the development of a Bulgarian TtS system, and
  • the development of a cultural web portal and the exploitation of the developed TtS system for enhancing it with speech, making it more accessible to visually impaired people

The above are complemented by diffusion activities for the dissemination of the R&D results of the project.