STFT

In addition to collecting the LTASS information I’m interested in the short-time Fourier transform (STFT). Which, is a Fourier-related transform used to determine the different sections of frequency and phase content in a signal. Basically, we divide the signal into segments of equal lengths and then compute the fast Fourier transform (FFT) of each segment yielding the Fourier spectrum as a function of time.

Removing Silence

For collecting LTASS AND STFT it was important to identify the silenced regions in speech because a lot of the CHILDES segments have silences and pauses. And by silence I mean “the absence of any signal characteristics, lowest energy compared to unvoiced and voiced speech segments, relatively more number of zero crossings compared to unvoiced segment and no correlation among successive samples.” Similarly, speech segmentation allows the identification of boundaries between words, syllables, or phonemes in spoken natural languages. It’s kinda like speeh recognition, except it’s goal is recognize non-verbal speech.

Best,
EO