IFAS-based voiced/unvoiced classification of speech signal
Dhany Arifianto, Takao Kobayashi · 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03). · 2003
This paper describes the application of the notion of instantaneous frequency amplitude spectrum (IFAS) to discriminate voiced and unvoiced segments of speech signals. The classification procedures of speech signals into voiced and unvoiced is determined by using a harmonicity measure acquired after evaluating the instantaneous frequency amplitude spectrum. Several voicing decisions based on harmonicity measure are described to provide possible modifications including a post-processing stage to improve the classifier accuracy. The performance of the proposed method is also compared against ESPS and AMDF to demonstrate the effectiveness of the proposed method.