This talk presents techniques for nonstationarity detection in the context of speech and audio waveforms, with broad application to any class of time series that exhibits locally stationary behavior.
