摘要
采用Fisher分类法实现清浊音判决.计算机模拟结果显示,该方法准确率达97.4%,优于传统的清浊音判决算法.
It is the key issue to improve the quality of speech processing.In this paper,a voicing decision algorithm using Fisher classification is discussed.Some rough achievements from computer simulation are presented.The correct rate for voicing decision of our method reaches 97.4%.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
1997年第5期30-32,共3页
Journal of Shanghai Jiaotong University
基金
国家攀登计划认知科学(神经网络)重大关键项目
国家自然科学基金
关键词
语音识别
语音编码
清浊音判决
Fisher分类法
speech recognition
speech coding
voicing decision
pitch extraction
pattern classification