基于增量调制序列的语音处理新方法

A NEW SPEECH SIGNAL PROCESSING METHOD BASED ON DELTA MODULATION SEQUENCE

本文提出了一个基于自适应增量调制序列(简称ADM序列)的语音分析处理方法。理论分析和实验结果均表明,这种新方法不仅在语音首尾判决,语音识别等方面具有良好的效果,同时比传统的方法易于用硬件实现,构成具有高性能价格比的语音识别系统,是一种具有商品化应用前景的新方法。

A new Speech Signal processing method based on detta modula- tion sequence is proposed in this paper.Theoretical analysis and experimental results show that the new method performs well in endpoints detection and speech recognition.Furthermore,the new algorithm can be realized by hardware cost effectively,thus having a promising commercial prospect.