Power-Weighted LPC Formant Estimation

A power-weighted formant frequency estimation procedure based on Linear Predictive Coding (LPC) is presented. It works by pre-emphasizing the dominant spectral components of an input signal, which allows a subsequent step to extract frequencies with greater accuracy. The accuracy traditional LPC improved this new estimator for different classes synthetic signals and speech. Power-weighted significantly reliably outperforms variants at task using VTR formants dataset, database consisting Vocal Tract Resonance (VTR) trajectories obtained human experts first three frequencies. This performance gain evident over range filter orders.