Bibliography

Next: About this document ... Up: A Computational Model of Previous: Conclusions

Bibliography

Bregman, 1990

A. S. Bregman. (1990). Auditory Scene Analysis: The Perceptual Organization of Sound, MIT Press, Cambridge, Mass.

Bregman, 1993

A. S. Bregman. (1993). ``Auditory Scene Analysis: hearing in complex environments,'' in Thinking in Sounds, (Eds. S. McAdams and E.Bigand), pp. 10-36, Oxford University Press, New York.

Moore, 1997

Brain C.J. Moore. (1997). An Introduction to the Psychology of Hearing, 4th ed., Academic Press, San Diego.

Moore, 1992

Brain C.J. Moore. (1992). ``Comodulation Masking release and Modulation Discrimination Interface,'' in The Auditory Processing of Speech, from Sound to Words, (Edited by M. E. H. Schouten), pp. 167-183, Mouton de Gruyter, NewYork.

Chui, 1992

C. K. Chui. (1992). An Introduction to Wavelets, Academic Press, Boston, MA.

Hall et al., 1984

Hall, J. W. and Fernandes, M. A. (1984). ``The role of monaural frequency selectivity in binaural analysis,'' J. Acoust. Soc. Am. 76, 435-439.

Hall et al., 1988

Hall, J. W. and Grose, J. H. (1988). ``Comodulation masking release: Evidence for multiple cues,'' J. Acoust. Soc. Am. 84, pp. 1669-1675.

Patterson et al., 1986

Patterson, R. D. and Moore, B. C. J. (1986). Auditory filters and excitation patterns as representations of frequency resolution. In Frequency Selectivity in Hearing (ed. B. C. J. Moore), Academic Press, London and New York.

Patterson et al., 1991

Patterson, R. D. and John Holdsworth. (1991). A Functional Model of Neural Activity Patterns and Auditory Images, Advances in speech, Hearing and Language Processing, vol. 3, JAI Press, London.

Unoki et al., 1997

Masashi Unoki and Masato Akagi. (1997). ``A Method of Signal Extraction from Noise-Added Signal,'' IEICE, Vol. J80-A, No. 3, pp. 444-453, March (in Japanese).

Willen et al., 1997

Willen A. C. van den Brink, Tammo Houtgast, and Guido F. Smoorenburg. (1992). ``Effectiveness of Comodulation Masking Release,'' in The Auditory Processing of Speech, from Sound to Words, (Eds. M. E. H. Schouten), pp. 167-183, Mouton de Gruyter, NewYork.

**Figure 1:** Computational model of CMR. This model consists of two models, our auditory segregation model (model A) and the power spectrum model of masking (model B), and a selection process that selects one of their results.
$\begin{figure} \begin{center} \vspace{5cm} \epsfile{file=FIGURE/CMRmodel.eps,width=0.95\textwidth} \end{center}\vspace{5cm} \end{figure}$

**Figure 2:** Model A: an auditory segregation model. This model consists of three parts: (a) an auditory filterbank, (b) separation block, and (c) grouping block.
$\begin{figure} \begin{center} \epsfile{file=FIGURE/ModelA.eps,width=0.95\textwidth} \end{center}\end{figure}$

**Figure 3:** Relationship between center frequency and ERB. Dashed-line shows ERB corresponding to the center frequency and solid-line shows linear approximation of ERB at 600 Hz.
$\begin{figure} \begin{center} \epsfile{file=FIGURE/ERB.eps,width=0.95\textwidth} \end{center}\end{figure}$

**Figure 4:** Segregation algorithm.
$% latex2html id marker 1624 \fbox{\footnotesize{ \begin{minipage}[h]{7cm} \begin... ...form) from Eqs. (\ref{eq:A_k}) and (\ref{eq:B_k}); \end{tabbing}\end{minipage}}}$

**Figure 5:** Bandpassed characteristics of a sinusoidal signal f₁(t) and bandpassed noise f₂(t) in the adjacent channel.
$\begin{figure} \begin{center} \epsfile{file=FIGURE/adjacent2.eps,width=0.95\textwidth} \end{center}\end{figure}$

**Figure 6:** Model B: a power spectrum model of masking.
$\begin{figure} \begin{center} \vspace{5cm} \epsfile{file=FIGURE/ModelB.eps,width=0.95\textwidth} \end{center}\vspace{5cm} \end{figure}$

**Figure 7:** Results of CMR (Hall *et al.*, 1984). The points labeled 'R' are thresholds for 1 kHz signal centered in a band of random noise, plotted as a function of the bandwidth of the noise. The points labeled 'M' are the thresholds obtained when the noise was amplitude modulated at an irregular, low rate.
$\begin{figure} \begin{center} \epsfile{file=FIGURE/Hall.eps,width=0.95\textwidth} \end{center}\end{figure}$

**Figure 8:** Stimuli: a sinusoidal signal f₁(t) (top), a bandpassed random noise f₂₁(t) (middle), and an AM bandpassed noise f₂₂(t) (bottom).
$\begin{figure} \epsfile{file=FIGURE/CMRdata.eps,width=0.95\textwidth} \par\epsfile{file=FIGURE/UCMRdata.eps,width=0.95\textwidth} \end{figure}$

**Figure 9:** Mixed signals f_M(t) (top) and f_R(t) (bottom).
$\begin{figure} \epsfile{file=FIGURE/Mixdata.eps,width=0.95\textwidth} \end{figure}$

**Figure:** Relationship between the bandwidth related to the number of adjacent auditory filters and the SNR for the extracted signal $\hat{f}_{1,A}(t)$ . The vertical and horizontal axes show the improved SNR of the extracted sinusoidal signal $\hat{f}_{1,A}(t)$ and the bandwidth related to L, respectively. The real line and the error bar show the mean and standard deviation of the SNR of the signal $\hat{f}_{1,A}(t)$ extracted from 300 mixed signals, respectively.
$\begin{figure} \begin{center} \epsfile{file=FIGURE/CMRa.eps,width=0.95\textwidth} \end{center}\end{figure}$

**Figure:** Relationship between the masker bandwidth and the SNR for the extracted signal $\hat{f}_{1,B}(t)$ . The vertical and horizontal axes show the improved SNR of the extracted sinusoidal signal $\hat{f}_{1,B}(t)$ and the bandwidth related to L, respectively. The real line and the error bar show the mean and standard deviation of the SNR of the signal $\hat{f}_{1,B}(t)$ extracted from 300 mixed signals, respectively.
$\begin{figure} \begin{center} \epsfile{file=FIGURE/CMRb.eps,width=0.95\textwidth} \end{center}\end{figure}$

**Figure 12:** Relationship between the masker bandwidth and the SNR for the extracted signal. This characteristic was obtained by the result of the selection process from Figs. 10 and 11.
$\begin{figure} \begin{center} \epsfile{file=FIGURE/CMR.eps,width=0.95\textwidth} \end{center}\end{figure}$

Next: About this document ... Up: A Computational Model of Previous: Conclusions

Masashi Unoki
2000-11-07