基于可视语音合成的3D通信技术研究

Research on 3D communication technologies based on text-to-visual speeches

提出了基于可视语音合成的3D通信技术概念。通信双方利用文本交换信息,用户终端采用可视语音合成技术对接收到的文字进行人物化朗读;同时通过3D用户替身的肢体动作和面部表情表达文字蕴涵的人类情感和意图信息,给出了基于可视语音合成的通信平台结构、情感和意图的表达模型、用户替身的自主交互模型。采用程序驱动方法,通过控制虚拟人各连杆间沿关节自由度方向夹角及角度变化率等特征参数实现肢体运动合成;利用VB6.0和OpenGL实现了基于可视语音合成的3D通信平台原型。

This paper proposed a concept of 3D communication technology based on text-to-visual speeches （TTVS）. Partners of a communication pair exchanged messages with text. At each end interface, adopted TTVS technology to read received text message. Simultaneously, a 3D avatar expressed embedded human emotions and intentions with body languanges and facial expressions. Proposed the structure of such a communication platform. Analyzed expression models of human emotions and intentions, and proposed autonomous interaction models of avatar. Driven by program, obtained avatar＇ s movement by controlling angles and their varying rates between neighbored links. Based on the proposed methods,developed a prototype of 3D communica- tion platform with function of TTVS under VB 6.0 and OpenGL.