改进LVAMDF及综合多因素基音检测算法

Improved Algorithm of Length Varied Average Magnitude Difference Function(LVAMDF) and Comprehensive Multi-factor for Pitch Frequency Detection

摘要:

在对语音信号静音、清音、浊音划分的基础上,针对语音信号周期特征明显段分布随机性问题,提出改进的变长度平均幅度差函数LVAMDF及综合多因素基音检测算法,该算法对语音信号进行周期特征明显段和周期特征不明显段的聚类划分,同时,获取周期特征明显语音段的所有基音周期的起止端点,针对少数基音周期划分倍频或半频问题,提出识别、修正方法,其识别、修正率极高；在对大量真实语音处理中,能够精确地检测出语音特征明显段的基音周期端点,基本没有倍频和半频划分,并且和AMDF、ACF算法作了对比。

Abstract:

On the basis of the methodology that categorizes speech signal into three types,silence,voiceless sound,voiced sound,in view of the random distribution of obvious periodic property speech,the improved algorithm of length varied average magnitude difference function (LVAMDF) and comprehensive multi-factor for pitch frequency detection is put forward to categorize voiced sound into two types,one is obvious periodic property speech,the other one is unobvious periodic property speech. At the same time,the starting and ending points of all accurate pitch period in the obvious periodic property speech is achieved. For a few pitch periods divided into frequency doubling or half frequency,the recognition and correction method is proposed which has a high recognition and correction rate. Finally,in a large amount of the real speech processing experiments,the method can detect the pitch period in the obvious periodic property speech accurately,and there is hardly frequency doubling and half frequency. The results show that the proposed algorithm performs much better on pitch detection compared with AMDF and ACF.