Fundamental frequency extraction by utilizing modified BaNa in noisy speech
Open Access
- 1 September 2024
- journal article
- Published by Institute of Advanced Engineering and Science in International Journal of Advances in Applied Sciences
- Vol. 13 (3) , 515-529
Abstract
A sound’s pitch can be largely understood and perceived by using its fundamental frequency. Multiple algorithms have been developed for extracting fundamental frequency, and the choice of which one to employ depends on the noise and features of the signal. Therefore, for an accurate fundamental frequency estimate, the noise resistance of the algorithm becomes even more crucial. Still, many of the most advanced algorithms fail to produce acceptable results when faced with loud speech recordings that have low signal-to-noise ratios (SNRs). In this research paper, we focus on the harmonic selection step in BaNa method, which is one of the vital parts for enhancing the extraction accuracy of fundamental frequency (F0) in noisy situations. BaNa algorithm always emphasizes 5 harmonics on average for both male and female speakers. However, our observation reveals that relying on 5 harmonics is inadequate for male speakers in noisy conditions. Thus, we propose a new idea based on BaNa that separately utilizes the 3 harmonics for male speakers and 5 harmonics for female speakers to achieve accurate pitch extraction within noisy environments. The results demonstrate that our proposed approach attains the lowest rate of gross pitch error (GPE) across various noise types and SNR levels.Keywords
This publication has 32 references indexed in Scilit:
- A Pitch Estimation Algorithm for Speech in Complex Noise Environments Based on the Radon TransformIEEE Access, 2023
- A Novel Pitch Detection Algorithm Based on Instantaneous Frequency for Clean and Noisy SpeechCircuits, Systems, and Signal Processing, 2022
- Assessment of pitch-adaptive front-end signal processing for children’s speech recognitionComputer Speech & Language, 2018
- A Comparative Analysis of Pitch Detection Methods Under the Influence of Different Noise ConditionsJournal of Voice, 2015
- Pitch tracking of acoustic signals based on average squared mean difference functionSignal, Image and Video Processing, 2008
- Digital Speech Transmission2006
- A modified cepstrum method for pitch extraction2002
- Weighted autocorrelation for pitch extraction of noisy speechIEEE Transactions on Speech and Audio Processing, 2001
- Assessment for automatic speech recognition: II. NOISEX-92: A database and an experiment to study the effect of additive noise on speech recognition systemsSpeech Communication, 1993
- Research of individuality features in speech waves and automatic speaker recognition techniquesSpeech Communication, 1986