{
"title": "An Approach to Solving a Permutation Problem of Frequency Domain Independent Component Analysis for Blind Source Separation of Speech Signals",
"authors": "Masaru Fujieda, Takahiro Murakami, Yoshihisa Ishida",
"country": null,
"institution": null,
"volume": "18",
"journal": "International Journal of Electrical, Computer, Energetic, Electronic and Communication Engineering",
"pagesStart": 1254,
"pagesEnd": 1259,
"ISSN": "1307-6892",
"URL": "http:\/\/waset.org\/publications\/3094",
"abstract": "Independent component analysis (ICA) in the\r\nfrequency domain is used for solving the problem of blind source\r\nseparation (BSS). However, this method has some problems. For\r\nexample, a general ICA algorithm cannot determine the permutation\r\nof signals which is important in the frequency domain ICA. In this\r\npaper, we propose an approach to the solution for a permutation\r\nproblem. The idea is to effectively combine two conventional\r\napproaches. This approach improves the signal separation\r\nperformance by exploiting features of the conventional approaches.\r\nWe show the simulation results using artificial data.",
"references": null,
"publisher": "World Academy of Science, Engineering and Technology",
"index": "International Science Index 18, 2008"
}