Cross-spectral methods for processing biological signals

We present methods for estimation of signal parameters and apply these methods to biological signals. The methods are based on cross-spectral phase which is computed from the phase of the short time Fourier transform. The methods are applied to acoustical biological signals, including human speech and dolphin sonar clicks. Specifically addressed are the problems of crisp narrowband time frequency representations from very small data sets, accurately estimating speech formants, blind recovery of the group delay of the transmission channel and equalization of time-frequency representations