International Conference on Acoustics, Speech, and Signal Processing · 1989 · 11 citations · 0 references
ComputingComputer ScienceEnglishComputational LinguisticsNatural Language ProcessingPhoneticsSpeech RecognitionSpeech SynthesisAcoustic AnalysisCross-lingual Speech SynthesisSpeech Recognition SoftwareSpeech Sound SynthesisSpeech Synthesis AcousticsSpeech-to-text SystemsText-to-speech SynthesisWord Recognition Accuracy
A MITalk-based [l] real-time system for unres tricted English, Japanese and Chinese text-tc-speech synthesis has been developed with high observed intelligibility (94% for word recognition in English and 70.8% for 119 Japanese monosyllables). Our approach in making the system multi-lingual has been simple yet effective: in developing the Japanese and Chinese systems, we have used the hardware and the program structure and tools of the system originally developed for English, but have otherwise kept the development of the different languages separate. In addition to substantial improvements and modifications of the original MITalk, two major innovations in our system are (1) the method for controlling the glottal pulse and (2) the Phoenix develop ment tools, which combine a powerful and flexible parameter editor with programs for acoustic analysis.