9
Comparative analysis of the ICA algorithms applied on a 2D signal
Authors: Nikola Bešić, Gabriel Vasile, Budimir Lutovac, Srdjan Stanković, Dragan Filipović
Number of views: 522
Blind source separation is one of the major areas of research in signal and image processing today. Being a broad area, it actually comprises various distinctive techniques. The revolutionary one, considered here, is Independent Component Analysis (ICA). Nowadays, the FastICA method, based on the fixed-point iterative algorithm, is probably the most widespread method in estimating independent components. Aside from it, there are numerous methods based on tensorial decomposition, among which FOBI and JADE are regarded as the most representative. In this paper we introduce these ICA methods and analyze their performances, illustrated through the application on a synthetic 2D signal (image).