| 摘要 |
A speech analyzer for extracting spectrum information and pitch information from natural speech wherein an accuracy of pitch extraction is enhanced by sampling pitch at a sampling frequency which is higher than a sampling frequency for analyzing the spectrum information. |
