论文信息 - Sound source separation via computational auditory scene analysis-enhanced beamforming

Sound source separation via computational auditory scene analysis-enhanced beamforming

The paper describes a new approach to sound source separation, called computational auditory scene analysis-enhanced beamforming (CASA-EB). It achieves increased signal separation performance by combining beamforming with an auditory model (CASA). The motivation for combining them is that they are complementary, i.e., they use independent signal attributes, and as a result, CASA-EB can separate more types of mixtures. The efficacy of this approach is demonstrated by experimental results.

[1] Guy J. Brown,et al. A comparison of auditory and blind separation techniques for speech segregation , 2001, IEEE Trans. Speech Audio Process..

[2] François Bourgeois,et al. An extension of the Munkres algorithm for the assignment problem to rectangular matrices , 1971, CACM.

[3] Woo Young Hong,et al. Focused Wide-Band Array Processing for Spatial Spectral Estimation , 1994 .

[4] Guy J. Brown,et al. Separation of speech from interfering sounds based on oscillatory correlation , 1999, IEEE Trans. Neural Networks.

[5] Daniel Patrick Whittlesey Ellis,et al. Prediction-driven computational auditory scene analysis , 1996 .

[6] Ali S. Hadi,et al. Finding Groups in Data: An Introduction to Chster Analysis , 1991 .

[7] P. N. Denbigh,et al. Pitch extraction and separation of overlapping speech , 1991, Speech Commun..

[8] Martin Cooke,et al. Modelling auditory processing and organisation , 1993, Distinguished dissertations in computer science.

[9] Tomohiro Nakatani,et al. Harmonic sound stream segregation using localization and its application to speech stream segregation , 1999, Speech Commun..