Sixth European Conference on Speech Communication and Technology
(EUROSPEECH'99)

Budapest, Hungary
September 5-9, 1999

Text to Speech Control Protocol

Jirí Hanika (1), Petr Horák (2)

(1) Faculty of Arts, Charles University, Prague, Czech Republic
(2) Institute of Radio Engineering and Electronics, Academy of Sciences, Prague, Chech Republic

The necessity of using the same TTS software within multiple operating environments, as well as separation of the TTS subsystem from the text source leads to a demand for a standard controlling interface. We describe our attempt at such standard controlling interface, the Text-to-Speech Control Protocol (TTSCP). The problems and the choices we have made for TTSCP as implemented within the framework of the recently introduced Epos Speech System are discussed. TTSCP is structured as a generic connection-oriented data stream processing protocol running atop the TCP protocol, and, while the control information exchange is human readable, it is tuned for efficient processing by server-side and both simple and complex client-side programs.


Full Paper (PDF)   Gnu-Zipped Postscript

Bibliographic reference.  Hanika, Jirí / Horák, Petr (1999): "Text to speech control protocol", In EUROSPEECH'99, 2143-2146.