作者: H. Hermansky , N. Morgan
DOI: 10.1109/89.326616
关键词: Stochastic resonance 、 Impulse response 、 Speech enhancement 、 Speech processing 、 Computer science 、 Noise 、 Convolution 、 Environmental noise 、 Speech recognition 、 Cepstral Mean and Variance Normalization
摘要: … We have been experimenting with filtering approaches that … to repeat the recognition experiment many times. To see … two control words (“yes” and “no”) were recorded at 8 kHz by …