论文部分内容阅读
文中推导了Bloomfield’s模型与LP模型的语音特征参数的关系,并导出了线性预测系数及相关的特征集。仿真结果表明,基于Bloomfield’s模型的语音信号建模是成功的,可以有效的进行语音信号预测及小词汇量孤立语音识别。
In this paper, the relationship between Bloomfield’s model and the speech feature parameters of the LP model is deduced, and the linear prediction coefficients and the related feature sets are derived. The simulation results show that the speech signal modeling based on Bloomfield’s model is successful and it can effectively predict the speech signal and isolate the speech with small vocabularies.