This source code is the implementation of the underdetermined convolutive blind source separation algorithm proposed in [1].

Quick Start
At first add the “ucbss_dist” folder to MATLAB Path.
Please run “main.m”. It will load 2 channel mixed speech signals in the “data” directory, and then it separates the mixed signals to 3 original source signals. The result files will be saved in the “result” directory.

How to Cite
If you find it helpful, please cite [1]

Contact Information
For any questions and comments, please send your email to bluehawk2k@kaist.ac.kr

[1] Janghoon Cho and Chang D. Yoo. "Underdetermined Convolutive BSS: Bayes Risk Minimization Based on a Mixture of Super-Gaussian Posterior Approximation." Audio, Speech, and Language Processing, IEEE/ACM Transactions on 23.5 (2015): 828-839.

Category Software /0
License GPL v2
Producer Admin
Package Regdate Nov 30, 2015 16:36 Package Downloaded 30
Latest version 1.0

Version History

Version 1.0
Downloaded Count 30
Voted /0
Registered Date Nov 30, 2015 16:36