基于汉语语音特征的多媒体同步方法  

Multimedia Synchronization Method Based on the Chinese Pronunciation Characters

作　　者：田友胜[1] 侯义斌[2] R-J. Beun 陈亚滨[1] 钱屹[1] 

机构地区：[1]西安交通大学电信学院,陕西西安710049 [2]北京工业大学,北京100022 [3]Eindhoven University of Technology

出　　处：《小型微型计算机系统》2004年第2期266-270,共5页Journal of Chinese Computer Systems

基　　金：国家教委跨世纪人才基金资助

摘　　要：现有一些同步方法在重复使用或丢弃媒体单元时没有选择性 ,媒体间同步粒度也比较大 .汉语语音由声母、韵母和声调组成 ,且韵母部分发音做适量调整不会改变汉字的语意 .本文结合汉语的语音特征提出了一种新的多媒体同步方法——基于汉语语音特征的多媒体同步 .该方法通过对汉字语音中韵母部分发音做适量的压缩或拉伸来达到多媒体同步的目的 ,具有对媒体单元选择性好、同步粒度小等优点 .仿真结果证明 ,该方法比原有的段首同步法有更好的同步效果 .At present, most of multimedia synchronization methods have no choices from the multimedia units to be duplicated or dropped, and synchronization granularities are bigger. Chinese pronunciation is composed by initial consonant, compound vowel and tone. Our experiments show that the words meaning will not be changed with a properly adjusting of the compound vowels pronunciation. By the character, we propose a multimedia synchronization method based on the Chinese pronunciation characters. Through properly adjusting the compound vowel's pronunciation, we can compress or prolong the pronunciation and the multimedia could be synchronized in this way. This multimedia synchronization method can select media units to adjust its pronunciation and its synchronization granularities are smaller. From the simulation result, it is shown that the proposed synchronization method is better than conventional synchronization methods.

关 键 词：多媒体 同步 语音 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]