Epoch extraction from emotional speech using single frequency filtering approach

Sudarsana Reddy Kadiri, B. Yegnanarayana

Abstract: Highlights•An approach, which exploits the nature of impulse-like excitation in the speech signal is explored.•Three properties of impulse are used to extract impulse-like discontinuities, namely: (a)An impulse in the time domain results in flat spectrum in the frequency domain.(b)An impulse event in the time domain is a high energy event, i.e., the strength of the impulse is substantially larger than the strengths of the samples in the neighborhood.(c)The effect of impulse is spread at all the frequencies.•We demonstrate the benefit of exploiting the impulse-like events in terms of epoch identification accuracy.•The results of the proposed methods for emotional speech is comparable to the results for neutral speech, and is better than the results from many of the standard methods.