Name
Abstract | ||
|---|---|---|
In multisensor signal processing (geophysics, underwater acoustic, etc.), the Singular Value Decomposition (SVD) is a useful tool to perform a separation of the initial dataset into two complementary subspaces. The SVD of the data matrix {x,i} provides two orthogonal matrices that convey information on propagation vectors and normalized wavelets. The constraint imposed by the orthogonality's condition for the propagation vectors introduce errors in the signal subspace. To relax this condition, another matrix of normalized wavelet is calculated exploiting the concept of Independent Component Analysis (ICA). Efficiency of this new separation tool using the combined SVD-ICA procedure is shown on realistic dataset. |
Year | Venue | Keywords |
|---|---|---|
2002 | Toulouse | independent component analysis,sensor fusion,singular value decomposition,source separation,vectors,svd-ica,data orthogonal matrix,multi sensor signal processing,noise subspace,normalized wavelet matrix,propagation vector,signal separation |
Field | DocType | ISSN |
Singular value decomposition,Signal processing,Orthogonal matrix,Pattern recognition,Matrix (mathematics),Orthogonality,Independent component analysis,Artificial intelligence,Signal subspace,Mathematics,Wavelet | Conference | 2219-5491 |
Citations | PageRank | References |
0 | 0.34 | 1 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Valeriu D. Vrabie | 1 | 0 | 0.34 |
| J.I. Mars | 2 | 161 | 14.94 |