Tracking of Multiple Fundamental Frequencies in Diplophonic Voices

Detection of extra pulses in synthesized glottal area waveforms of dysphonic voices

“…5 . The method is adapted from the one described in [ 14 ]. A 32 ms Hann window with a 16 ms overlap is used for blocking signals.…”

“…A 32 ms Hann window with a 16 ms overlap is used for blocking signals. Candidate f o -tracks foγ are obtained by picking peaks in the spectrum of the GAW d ′( n ), and applying the Viterbi algorithm six times, as in the “fast” setup described in [14]. The candidate index γ = 1, 2, …, Γ , and Γ is the number of candidates.…”

“…The candidate index γ = 1, 2, …, Γ , and Γ is the number of candidates. No high-pass filtering is used, as was for the analysis of audio signals in [14]. Candidate cyclic unit pulse trains u1γ(n) are created for each foγ. Candidate cyclic pulse shapes r γ ( l ) are obtained by cross-correlating candidate u1γ(n) with the observed GAW d ′( n ).…”

“…Candidate cyclic unit pulse trains u1γ(n) are created for each foγ. Candidate cyclic pulse shapes r γ ( l ) are obtained by cross-correlating candidate u1γ(n) with the observed GAW d ′( n ). The candidate f o -tracks foγ and the pulse shapes’ discrete Fourier coefficients a γ and b γ are used in a Fourier synthesizer, which determines candidate cyclic pulse trains d1γ(n). For further details the interested reader is referred to [14].…”

“…We propose “ultra fast” candidate selection that replaces the candidate selection approach described in [14]. The estimate of the cyclic pulse train d 1 ( n ) is given by trued^1(n)=∑γ=1Γsγ⋅d1γ(n), where the binary candidate selection vector S = s γ ∈ {0, 1}, and Γ is the number of candidates.…”