2018
DOI: 10.14209/jcis.2018.16
|View full text |Cite
|
Sign up to set email alerts
|

On the Darmois-Skitovich Theorem and Spatial Independence in Blind Source Separation

Abstract: In many signal processing applications, one may come across the need to individually recover unobserved signals which are also combined in an unknown manner. This problem is widely known as blind source separation (BSS). One of the most prominent set of BSS techniques, known as independent component analysis (ICA), owes much of its development and theoretical understanding to the Darmois-Skitovich theorem. Although this theorem is implicitly used in BSS to establish source separability conditions in ICA, littl… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
1
0
3

Year Published

2019
2019
2022
2022

Publication Types

Select...
1
1

Relationship

1
1

Authors

Journals

citations
Cited by 2 publications
(4 citation statements)
references
References 19 publications
(44 reference statements)
0
1
0
3
Order By: Relevance
“…This lemma has been initially introduced to prove the Darmois-Skitovich theorem, and is generally employed to analyze the identification in static ICA (see e.g. Comon, (1994, Lemma 20;Pavan and Miranda, 2018, Lemma 1).…”
Section: Data Availability Statementmentioning
confidence: 99%
“…This lemma has been initially introduced to prove the Darmois-Skitovich theorem, and is generally employed to analyze the identification in static ICA (see e.g. Comon, (1994, Lemma 20;Pavan and Miranda, 2018, Lemma 1).…”
Section: Data Availability Statementmentioning
confidence: 99%
“…Essas ambiguidades são previstas pelo teorema de Darmois-Skitovich e decorrentes de limitações teóricas desse problema particular de separação cega [3], [19].…”
Section: Aplicação a Um Caso De Separação Cegaunclassified
“…Observando-se (10), nota-se que se K s i = 0 para mais de uma fonte, então U nunca será identificável. Felizmente, exceto pelo caso gaussiano que pode ser desconsiderado por limitações teóricas [19], distribuições com autocumulante de ordem quatro nulo são pouco usuais [13]. Excluído esse caso, resta saber como garantir que Q x (V ) tenha autovalores distintos sempre, independentemente das estatísticas das fontes e das colunas de U .…”
Section: Aplicação a Um Caso De Separação Cegaunclassified
See 1 more Smart Citation