Results and considerations

The segregation accuracy of the three simulations and the four comparisons is shown in Fig. . In this figure, the bar height shows the mean of segregation accuracy and the error bar shows the standard deviation of segregation accuracy. The results show that the segregation accuracy using the proposed model was better than that using the other three methods. These results show that the proposed model can segregate the desired vowel from a noisy vowel precisely even in waveforms. In addition, the result of the comparison between the proposed model and (2) shows that the simultaneous signals can be precisely segregated using the instantaneous amplitude and phase. As a result of comparison between the proposed method and (3), improvements in segregation accuracies at the SNR of 5 dB for simulations 1, 2, and 3 are about 10 dB, about 9 dB, and about 7 dB, respectively.

  

Figure: Segregation accuracies for simulations. (a) AM-FM complex tone, (b) vowel, (c) continuous vowel.