Robust pitch tracking for prosodic modeling in telephone speech

In this paper, we introduce a pitch detection algorithm that is particularly robust for telephone speech and prosodic modeling. The algorithm uses a logarithmically sampled spectral representation of speech, similar to that in the subharmonic summation approach [2]. Constraints for log F0 and ∆ logF0 are combined in a dynamic programming search to find an optimum pitch track. The search algorithm is able to find a continuous pitch contour regardless of the voicing status, while a separate voicing decision module computes a probability of voicing per frame. We evaluated the algorithm using the Keele pitch extraction reference database [4] under both studio and telephone conditions. Our algorithm is very robust to channel degradation, and compares favorably to xwaves under telephone conditions. It also significantly outperforms xwaves when used for tone classification on a telephone quality Mandarin digit corpus.