Waveform to Single Sinusoid Regression to Estimate the F0 Contour from Noisy Speech Using Recurrent Deep Neural Networks.

Akihiro Kato, Tomi Kinnunen

Browse the full Interspeech paper archive.