A spectral/temporal method for robust fundamental frequency tracking

Abstract

In this paper, a fundamental frequency (F(0)) tracking algorithm is presented that is extremely robust for both high quality and telephone speech, at signal to noise ratios ranging from clean speech to very noisy speech. The algorithm is named "YAAPT," for "yet another algorithm for pitch tracking." The algorithm is based on a combination of time domain processing, using the normalized cross correlation, and frequency domain processing. Major steps include processing of the original acoustic signal and a nonlinearly processed version of the signal, the use of a new method for computing a modified autocorrelation function that incorporates information from multiple spectral harmonic peaks, peak picking to select multiple F(0) candidates and associated figures of merit, and extensive use of dynamic programming to find the "best" track among the multiple F(0) candidates. The algorithm was evaluated by using three databases and compared to three other published F(0) tracking algorithms by using both high quality and telephone speech for various noise conditions. For clean speech, the error rates obtained are comparable to those obtained with the best results reported for any other algorithm; for noisy telephone speech, the error rates obtained are lower than those obtained with other methods.

Decision Support System for Keyword Spotting Using Theory of Evidence

“…This approach deploys a pitch tracking method proposed in [16] to estimate the fundamental frequency using multiple information resources. The enhanced speech signals are then forwarded for the feature extraction process to be expressed as a sequence of feature vectors that may provide sufficient information to represent the speech utterance.…”

Acoustic-Phonetic Feature Based Dialect Identification in Hindi Speech

“…Several algorithm for F0 estimation has been proposed in literature and can be broadly categorized based on their feature's domain i.e. time domain, frequency domain, hybrid time and frequency domain and event detection methods [29][30][31]. In this work,…”

“…Several studies exist to show the importance of prosodic features such as word accent and the phrase intonation in human speech processing, but very few use F0 in combination with other acoustic features [27]. F0 has been mostly found suitable for tonal languages, but the presence of prosodic tone due to accent influenced by native dialect of speakers [26,29,30] motivated us to investigate it for Hindi dialect recognition.…”

“…YAAPT ("Yet Another Algorithm for Pitch Tracking") [30], a noise robust and fast algorithm has been used. This algorithm works in combination of time domain and frequency domain processing and produces F0 value for each frame.…”

Robustness against distortion of fundamental frequency cues in simulated electro-acoustic hearing

“…The calculation of the fundamental frequency was performed with the YAAPT algorithm described in Zahorian and Hu (2008) because of its accuracy (see Zahorian and Hu, 2008).…”