摘要
针对强噪声背景下语音基音检测这一困难的问题,提出了一种基于计算声场景分析的基音检测的改进算法。该方法利用人的听觉感知特性,引入广义自相关函数以增强基音信息,适用于低信噪比和存在其它语音干扰下基音信息的提取。另外,对于混叠语音的基音周期轨迹的检测,提出一种比较简单的方法。实验结果证明,提出的算法是十分有效的。
Based on computational auditory scene analysis（CASA）, an improved pitch detection algorithm is proposed to solve the problem of pitch detecting in strong noisy background. In this algorithm, human auditory feature is utilized, and the generalized autocorrelation function is introduced to strengthen the pitch information Above-mentioned method could be applied to pitch detection in low SNR and in other speech disturbance. Then, a simple method is proposed to detect the pitch of mixed speech. The simulation experiments show that the algorithm is very efficient and effective.
出处
《通信技术》
2009年第12期164-166,169,共4页
Communications Technology
关键词
计算声场景分析
基音检测
广义自相关函数
computational auditory scene analysis
pitch detection
generalized autocorrelation function
