Design and implementation of Smart Voice system

Aiming at the problem that the current speech synthesis systems has strong taste of speech machine and poor voice quality,Smart Voice,a kind of Chinese TTS system was designed,and the Smart Voice call implementation was given.The system used a hierarchical,modular program structure,completing the marker analysis,syntax analysis,sustain treatment,tone sandhi processing function.Especially in the sustain processor,waveform concatenation technology innovation was used,creating waveform internal circulation damper module to make a good degree of natural sustain.System operation results showed that Smart Voice sounded natural,and could provide high-speed flexible Chinese speech for the network,games and other computer applications.