三维语音动画聊天室的设计与实现  

The Design and Implementation of a3D Speech Animation Chat-room

作　　者：王洵[1] 张道义[1] 董兰芳[1] 万寿红[1] 

机构地区：[1]中国科学技术大学计算机科学技术系,合肥230027

出　　处：《计算机工程与应用》2004年第1期106-108,共3页Computer Engineering and Applications

基　　金：安徽省自然科学基金(编号:01042203)

摘　　要：聊天室是人们在网上交流的一种重要手段,由于硬件设备、网络带宽的限制,目前广泛使用的聊天室只能基于文本和语音,不能基于人脸形象。以已经实现的“一个基于SAPI5.0的中文语音动画系统”为基础,设计并实现了一个结合文本、语音和人脸动画的三维语音动画聊天室。聊天室由客户端和服务器两部分组成,多个用户利用客户端连接到服务器上,用户可以输入文本并且选择各种表情,由客户端混合成带有表情标签的文本传送给服务器。服务器将该用户的三维人脸模型以及带有表情标签的文本传送给接收方,由接收方的客户端合成为带有表情的语音动画。聊天室只有发送方第一次给接收方传送信息时才需要下载人脸模型,以后只传送带有表情标签的文本,具有计算方法简单、系统通信开销小的特点,在普通微机上就能够产生带有表情的高质量的语音动画。Ch at-room is an important means that people use to communicate with each other i n Internet.Because of the limitation of hardware device and network bandwidth, people can merely use text and speech and can't use their faces in state-of-ar t chat-rooms.Based on the SAPI5.0-based Chinese speech animation system we have developed,we design and implement a3D speech animation chat-room whi ch combines text,speech and facial animation.The chat-room consists of a ser ver and some clients.A user can use a client to connect to the server,input some text and choose a kind of expression.The client sends text and expressi on to the server.The server transmits text and expression,accompanied by the s ender's face model,to the receiver.The client of the receiver uses these data to synthesize speech animation with expression.The server downloads the sender 's face model to the receiver only when it's the first time the sender sends me ssage to the receiver.Afterwards,the server does nothing but transmitting tex t and expression.Because the computing method is simple and the communication c ost is small,the chat-room can produce high-quality speech animation with exp ression in an ordinary microcomputer.

关 键 词：语音动画 聊天室 可视语音合成 ACTIVEX WINSOCK 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]