摘要
LPC美尔倒频谱系数(LPCMCC)根据人耳听觉特性将LPC倒谱系数(LPCC)用非线性美尔尺度进行变换,从而更接近人耳的听觉系统。基于此提出一种采用LPCMCC为特征参数,通过动态时间规整算法进行音频数据的分类与检索方法。实验结果表明,该方法的检索精度高于LPCC,检索效率高于美尔倒谱系数(MFCC)。
LPC Mel Cepstrum Coefficient(LPCMCC) which comes from LPC Cepstrum(LPCC) changed nonlinear by Mel scale according to human auditory characteristics, thus closer to human auditory system. This paper presents a method based on the feature parameter of LPCMCC, and through using dynamic time warping algorithm to audio data classificate and retrieve. Experimental results show that the accuracy rate of this method is better than LPCC, and the efficiency of retrieval is significantly higher than Mel Frequency Cepstrum Coefflcients(MFCC).
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第11期246-247,253,共3页
Computer Engineering
基金
福建省自然科学基金资助项目(A0510006)
关键词
LPC倒谱系数
LPC美尔倒谱系数
音频检索
美尔倒谱系数
LPC Cepstrum(LPCC)
LPC Mel Cepstrum Coefficient(LPCMCC)
audio retrieval
Mel Frequency Cepstrum Coefficients(MFCC)