Independent Component Analysis of Signals using Local Exponential Nonlinearities
スポンサーリンク
概要
- 論文の詳細を見る
In this paper we propose exponential type nonlinearities in order to blindly separate instantaneous mixtures of signals with symmetric probability distributions using the online relative gradient algorithm. These nonlinear functions are applied only in a certain range around zero in order to ensure the stability of the separating algorithm. The proposed truncated nonlinearities neutralize the effect of outliers while the higher order terms inherently present in the exponential function result in fast convergence especially for signals with bounded support. By varying the truncation threshold, signals with both sub-Gaussian and super-Gaussian probability distributions can be separated. For certain class of probability distributions (generalized Gaussian model), the optimal size of the threshold is obtained by examining the local stability conditions of the relative gradient algorithm. In order to separate sources consisting of both sub-Gaussian and super-Gaussian signals, we chose an adequate value of the threshold parameter based on the sign of normalized kurtosis estimated from the observed data. Finally, some computer simulations are presented to demonstrate the superior performance of the proposed idea.
- 社団法人電子情報通信学会の論文
- 2005-06-20
著者
-
Tufail Muhammad
Graduate School Of Engineering Tohoku University
-
Abe Masahide
Graduate School Of Engineering At Tohoku University
-
KAWAMATA Masayuki
Graduate School of Engineering, Tohoku University
-
Kawamata Masayuki
Graduate School Of Engineering Tohoku University
関連論文
- Roundoff Error Analysis in the Decoding of Fractal Image Coding Using a Simplified State-Space Model (Special Section of Papers Selected from ITC-CSCC '98)
- Analysis of Scaling-Factor-Quantization Error in Fractal Image Coding
- State-Space Approach to Roundoff Error Analysis of Fractal Image Coding
- An Extension to the Natural Gradient Algorithm for Robust Independent Component Analysis in the Presence of Outliers(Digital Signal Processing)
- A Characteristic Function Based Contrast Function for Blind Extraction of Statistically Independent Signals(Papers Selected from the 20th Symposium on Signal Processing)
- Independent Component Analysis of Signals using Local Exponential Nonlinearities
- Independent Component Analysis of Signals using Local Exponential Nonlinearities
- Independent Component Analysis of Signals using Local Exponential Nonlinearities
- Independent Component Analysis of Signals using Local Exponential Nonlinearities
- Evolutionary Digital Filtering Based on the Cloning and Mating Reproduction (Special Section of Selected Papers from the 8th Karuizawa Workshop on Circuits and Systems)
- A Technique for Active Noise Control Systems with Online Secondary Path Modeling Using Additive Noise Power Scheduling
- A Technique for Active Noise Control Systems with Online Secondary Path Modeling Using Additive Noise Power Scheduling
- A Technique for Active Noise Control Systems with Online Secondary Path Modeling Using Additive Noise Power Scheduling
- FOREWORD
- Parallel Implementation of Evolutionary Digital Filters in ASIC
- An Encoding Algorithm for IFS Coding of Homogeneous Fractal Images Using Univariate Polynomial Manipulation (Special Section on Digital Signal Processing)