Independent Component Analysis and Blind Source Separation

被引:12
作者
Barros, Allan Kardec
Principe, Jose Carlos
Erdogmus, Deniz
机构
[1] Univ Florida, Gainesville, FL USA
[2] Oregon Hlth & Sci Univ, Portland, OR 97201 USA
关键词
D O I
10.1016/j.sigpro.2007.03.013
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
The sixth international workshop on independent component analysis (ICA)/blind source separation (BSS) was held in Charleston, South Carolina, USA, in march 2006. About 122 papers from 18 countries focused on the use of independent component analysis (ICA) and bind source separation (BSS), in field of sensory and perceptual processing, were presented. Davies and James presented result on the separability conditions of sources for single-channel blind source separation. Araki et al. proposed features of different sources in an underdetermined measurement setting and demonstrated source-feature classification for proper separation of speech waveforms with three sensors positioned as a nonlinear array. Vincent et al presented different types of performance bound estimators for blind source separation such as time-invariant filtering and time-frequency masking. Di Persia et al. presented correlation between various quality of measures of blind speech separation.
引用
收藏
页码:1817 / 1818
页数:2
相关论文
empty
未找到相关数据