Frequency-Warping Invariant Features ITG-Fachtagung Sprachkommunikation 2006 Auditory Filterbank Based Frequency-Warping Invariant Features for Automatic Speech Recognition
Frequency-Warping Invariant Features ITG-Fachtagung Sprachkommunikation 2006 Auditory Filterbank Based Frequency-Warping Invariant Features for Automatic Speech Recognition
复制标题
频率扭曲不变特征 ITG-Fachtagung Sprachkommunikation 2006 用于自动语音识别的基于听觉滤波器组的频率扭曲不变特征
DOI：
--
复制
发表时间：
2006
期刊：
影响因子：
--
通讯作者：
Jan Rademacher
中科院分区：
文献类型：
--
作者：
Jan Rademacher
听觉滤波器组在自动语音识别系统的预处理阶段有很长的历史，最突出的例子是梅尔频率倒谱系数（MFCC）。在本文中，我们研究的有用性，滤波器组分析作为一个预处理器产生的频率扭曲不变的功能。结果表明，伽玛滤波器组分析后的等效矩形带宽（ERB）规模产生最强大的功能集。当训练集和测试集中的声道长度不同时，性能的改善是最显著的，这在例如儿童语音要用主要在成人数据上训练的系统来识别时是重要的。
Auditory filterbanks have a long history in the preprocessing stage of automatic speech recognition systems, with the most prominent examples being the mel frequency cepstral coefficients (MFCCs). In this paper, we study the usefulness of auditory-filterbank analyses as a preprocessor for the generation of frequency-warping invariant features. The results indicate, that gammatone-filterbank analyses following the equivalent rectangular bandwidth (ERB) scale yield the most robust feature sets. The performance improvements are most significant when the vocal tract lengths in the training and test sets differ, which is important when, for example, children speech is to be recognized with a system that was mainly trained on adult data.