论文部分内容阅读
提出使用肌电信号的语音识别系统。研究证实从面部肌肉中提取的肌电信号存在语音信息。实验使用(0～9)十个数字,受试者每隔10s重复单词。讲话时用电极记录五个通道表面肌电信号。用短时傅里叶变换提取信号的特征量,并通过主成分分析降维,有效地提取特征量进行模式识别。分类错误范围在15%以下。实验表明表面肌电信号的语音识别系统有着极好的前景。
Proposed the use of EMG voice recognition system. Research confirms the existence of voice messages from EMG extracted from facial muscles. The experiment used (0 ~ 9) ten digits, subjects repeated words every 10s. When speaking, record five channel surface EMG signals with electrodes. The feature of the signal is extracted by short-time Fourier transform, and the principal component analysis is used to reduce the dimension, and the feature quantity is effectively extracted for pattern recognition. Category error range below 15%. Experiments show that the surface EMG speech recognition system has a very good prospect.