Time delay estimation and adaptive frame length iterations for noise robust pitch extraction

We describe a pitch tracking method based on a time delay estimation technique. Given two frames of voiced speech and it’s time delay, we will run a linear regression on the unwrapped phase of the quotient of the spectrum of both frames. A weighted linear regression will allow us to avoid the effect of phases corrupted by spectral leakage and noise. Iterations adapting the frame length will allow us to have a better time resolution, avoiding inaccuracies that could come from pitch doubling and jitter. Key-Words: Pitch detection, noise robustness, time delay estimation.