An improved sparse component analysis (SCA) is developped. The SCA method is just defined in a framework before, but there no existing complete algorithm. We explore a compelte and automatical algorithm, then use it to deal with modal identification issue in machnical engineering. This software is just suitable for vibration signals, not for speech signal. If you want to process speech signals, you need to change the mixing matrix estimation method.
Dear Pro. Ishwarya Venkatesh, the corresponding paper has been submitted to shock and vibration.
Due to SCA based on instaneous mixing model, so it is only able to process sensor data without time-delay. So I advice you processing the data recorded in rigid structure instead of flexible structure.
Cite As
YuGang (2021). Sparse blind source separation,Sparse component analysis (https://www.mathworks.com/matlabcentral/fileexchange/48641-sparse-blind-source-separation-sparse-component-analysis), MATLAB Central File Exchange.
Retrieved March 18, 2021.
Comments and Ratings (8)
2 May 2019
29 Oct 2018
Excellent Work
2 Aug 2018
你好,关于这个程序,有相关的发表的文章可以参考下吗?非常感谢
5 Apr 2018
24 Sep 2017
您好,请问代码是不是缺少一个combntns函数啊,无法成功运行。。。
28 Oct 2015
@luo yisheng
这个代码是必须至少两通道数据的。不过你可以利用emd或者小波包之类的,将单通道分解为多通道,而后带入sca即可。
28 Oct 2015
您好!我想请问一下您的SCA代码能用于单通道盲源分离么?谢谢!
30 Mar 2015
你好,我用了你的代码,打算引用你这篇文章,希望能联系到你,我的邮箱1936390177@qq.com