Title :
Fine pitch contour extraction by voice fundamental wave filtering method
Author_Institution :
Electrotech. Lab., Ibaraki, Japan
Abstract :
A new method for fine pitch contour extraction is presented. The primary advantage of our approach is that a pitch period is determined pitch-synchronously using a simple wave filtering technique to extract the fundamental frequency component of voice. The first stage of processing is wave filtering using an adaptively controlled filter to extract the pitch channel signal. The second stage is a pitch contour determination process of which there are two different approaches. The first involves segmenting the wave into pitch periods by detecting zero crossing points. The other involves extracting continuous pitch and intensity contours. A performance study on the algorithm showed very good agreement between the resultant pitch contours obtained with this method and the manually traced standard pitch contours
Keywords :
adaptive filters; feature extraction; filtering theory; speech processing; adaptively controlled filter; algorithm; feature extraction; fine pitch contour extraction; intensity contours; performance; pitch channel signal; pitch contour; pitch period; speech processing; voice fundamental wave filtering method; wave segmentation; zero crossing points detection; Adaptive filters; Band pass filters; Data mining; Equations; Filtering; Frequency; Power harmonic filters; Signal processing; Signal processing algorithms; Speech;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1994. ICASSP-94., 1994 IEEE International Conference on
Conference_Location :
Adelaide, SA
Print_ISBN :
0-7803-1775-0
DOI :
10.1109/ICASSP.1994.389687