Informed FastICA: Semi-Blind Minimum Variance Distortionless Beamformer
Abstract
Non-Gaussianity-based Independent Vector Extraction leads to the famous one-unit FastICA/FastIVA algorithm when the likelihood function is optimized using an approximate Newton-Raphson algorithm under the orthogonality constraint. In this paper, we replace the constraint with the analytic form of the minimum variance distortionless beamformer (MVDR), by which a semi-blind variant of FastICA/FastIVA is obtained. The side information here is provided by a weighted covariance matrix replacing the noise covariance matrix, the estimation of which is a frequent goal of neural beamformers. The algorithm thus provides an intuitive connection between model-based blind extraction and learning-based extraction. The algorithm is tested in simulations and speaker ID-guided speaker extraction, showing fast convergence and promising performance.
Keywords
Cite
@article{arxiv.2407.09259,
title = {Informed FastICA: Semi-Blind Minimum Variance Distortionless Beamformer},
author = {Zbyněk Koldovský and Jiří Málek and Jaroslav Čmejla and Stephen O'Regan},
journal= {arXiv preprint arXiv:2407.09259},
year = {2024}
}
Comments
accepted for IWAENC 2024