-
Something wrong with this record ?
A hybrid technique for blind separation of non-gaussian and time-correlated sources using a multicomponent approach
P Tichavsky, Z Koldovsky, A Yeredor, G Gomez-Herrero, E Doron
Language English Country United States
- MeSH
- Algorithms MeSH
- Time MeSH
- Financing, Organized MeSH
- Humans MeSH
- Neural Networks, Computer MeSH
- Signal Processing, Computer-Assisted MeSH
- Models, Statistical MeSH
- Check Tag
- Humans MeSH
Blind inversion of a linear and instantaneous mixture of source signals is a problem often encountered in many signal processing applications. Efficient fastICA (EFICA) offers an asymptotically optimal solution to this problem when all of the sources obey a generalized Gaussian distribution, at most one of them is Gaussian, and each is independent and identically distributed (i.i.d.) in time. Likewise, weights-adjusted second-order blind identification (WASOBI) is asymptotically optimal when all the sources are Gaussian and can be modeled as autoregressive (AR) processes with distinct spectra. Nevertheless, real-life mixtures are likely to contain both Gaussian AR and non-Gaussian i.i.d. sources, rendering WASOBI and EFICA severely suboptimal. In this paper, we propose a novel scheme for combining the strengths of EFICA and WASOBI in order to deal with such hybrid mixtures. Simulations show that our approach outperforms competing algorithms designed for separating similar mixtures.
References provided by Crossref.org
- 000
- 02322naa 2200325 a 4500
- 001
- bmc11003033
- 003
- CZ-PrNML
- 005
- 20121105124422.0
- 008
- 110225s2008 xxu e eng||
- 009
- AR
- 024 __
- $a 10.1109/tnn.2007.908648 $2 doi
- 035 __
- $a (PubMed)18334362
- 040 __
- $a ABA008 $b cze $c ABA008 $d ABA008 $e AACR2
- 041 0_
- $a eng
- 044 __
- $a xxu
- 100 1_
- $a Tichavský, Petr. $7 _AN059680
- 245 12
- $a A hybrid technique for blind separation of non-gaussian and time-correlated sources using a multicomponent approach / $c P Tichavsky, Z Koldovsky, A Yeredor, G Gomez-Herrero, E Doron
- 314 __
- $a Institute of Information Theory and Automation, Academy of Sciences of the Czech Republic, 182 08 Prague8, Czech Republic. tichavsk@utia.cas.cz
- 520 9_
- $a Blind inversion of a linear and instantaneous mixture of source signals is a problem often encountered in many signal processing applications. Efficient fastICA (EFICA) offers an asymptotically optimal solution to this problem when all of the sources obey a generalized Gaussian distribution, at most one of them is Gaussian, and each is independent and identically distributed (i.i.d.) in time. Likewise, weights-adjusted second-order blind identification (WASOBI) is asymptotically optimal when all the sources are Gaussian and can be modeled as autoregressive (AR) processes with distinct spectra. Nevertheless, real-life mixtures are likely to contain both Gaussian AR and non-Gaussian i.i.d. sources, rendering WASOBI and EFICA severely suboptimal. In this paper, we propose a novel scheme for combining the strengths of EFICA and WASOBI in order to deal with such hybrid mixtures. Simulations show that our approach outperforms competing algorithms designed for separating similar mixtures.
- 650 _2
- $a algoritmy $7 D000465
- 650 _2
- $a lidé $7 D006801
- 650 _2
- $a statistické modely $7 D015233
- 650 _2
- $a neuronové sítě $7 D016571
- 650 _2
- $a počítačové zpracování signálu $7 D012815
- 650 _2
- $a čas $7 D013995
- 650 _2
- $a financování organizované $7 D005381
- 700 1_
- $a Koldovský, Zbyněk. $7 _AN059681
- 700 1_
- $a Yeredor, Arie
- 700 1_
- $a Gomez-Herrero, German
- 700 1_
- $a Doron, Eran
- 773 0_
- $t IEEE Transactions on Neural Networks $w MED00002175 $g Roč. 19, č. 3 (2008), s. 421-430 $x 1045-9227
- 910 __
- $a ABA008 $b x $y 1
- 990 __
- $a 20110413114032 $b ABA008
- 991 __
- $a 20121105124429 $b ABA008
- 999 __
- $a ok $b bmc $g 830433 $s 695025
- BAS __
- $a 3
- BMC __
- $a 2008 $b 19 $c 3 $d 421-430 $i 1045-9227 $m IEEE transactions on neural networks $n IEEE Trans Neural Netw $x MED00002175
- LZP __
- $a 2011-2B/dkme
