Figs. 6, 8 and 10 show the auditory-oriented spectral distortion (ASDs) in the three simulations. In these figures, the bar height shows the mean of the ASD and the error bar shows its standard deviation. Figs. 7, 9 and 11 show the segregation accuracies in the three simulations. In these figures, the bar height shows the mean of the segregation accuracy and the error bar shows its standard deviation.
The results show that the proposed method is better than our previous model, and that it obtained better segregation accuracy than the other five methods. The proposed model can precisely segregate a desired vowel from a noisy vowel even in waveforms, and it can reduce the ASD for sound segregation. However, we cannot conclude that it can precisely segregate a desired vowel form a noisy vowel in hearing, by using only the ASD.] We need to do hearing tests by using a subjective measure. The comparisons with conditions (2) and (3) show that the simultaneous signals can be precisely segregated using the instantaneous amplitude and phase. The comparison with condition (4) shows that the proposed model is more useful than spectral subtraction in the measure of segregation accuracy and the ASD. The comparison with condition (5) shows that the improvements in segregation accuracies at an SNR of 5 dB for simulations 1, 2 (in the case of pink-noise), and 3 were about 12, 8, and 5 dB, respectively.