Comparison-of-Blind-Source-Separation-techniques

This sample software applies Blind Source Separation (BSS) algorithms to simulated convolutive mixtures in order to compare separation performance.

1 - Offline BSS comparison

The main file to run comparison: "main_offline_bss_separation.m": AIRES vs TRINICON vs ILRMA vs AuxIVA

The main executable file separates convolutive mixture using our proposed Time Domain Stereo Audio Source Separation technique (AIRES)[1], time domain TRINICON[2], frequency domain BSS ILRMA[3,4] and AuxIVA[5]. For evaluation the Signal-to-Distortion Ratio, Signal-to-Interference Ratio and computation time for all 4 BSS are printed.

How to change convolutive mixtures:

simply uncomment in the beginning of the file desired line with FileName "fname".

Pay attention - all BSS work only with audio signals of 16kHz sampling rate.

[1] - O. Golokolenko and G. Schuller, "Fast time domain stereo audio source separation using fractional delay filters," in AES 147th Convention, 2019.

[2] - H. Buchner, R. Aichner, and W. Kellermann, "Trinicon: A versatile framework for multichannel blind signal processing," in IEEE International Conference on Acoustics, Speech, and Signal Processing, Montreal, Que.,Canada, 2004.

[3] - D. Kitamura, N. Ono, H. Sawada, H. Kameoka, H. Saruwatari, "Determined blind source separation unifying independent vector analysis and nonnegative matrix factorization," IEEE/ACM Trans. ASLP, vol. 24, no. 9, pp. 1626-1641, September 2016.

[4] - D. Kitamura, N. Ono, H. Sawada, H. Kameoka, H. Saruwatari, "Determined blind source separation with independent low-rank matrix analysis," Audio Source Separation. Signals and Communication Technology., S. Makino, Ed. Springer, Cham, pp. 125-155, March 2018.

[5] - Ono, Nobutaka. "Stable and fast update rules for independent vector analysis based on auxiliary function technique." WASPAA 2011.

2 - Online BSS comparison

The main file to run comparison: "main_online_bss_separation.m": AIRES vs TRINICON

The main executable file separates convolutive mixture using our proposed Time Domain Stereo Audio Source Separation technique (AIRES)[1], online time domain TRINICON[2]. For evaluation the Signal-to-Distortion Ratio, Signal-to-Interference Ratio and mean computation time per signal block are printed.

How to change convolutive mixtures:

simply uncomment in the beginning of the file desired line with FileName "fname".

Pay attention - all BSS work only with audio signals of 16kHz sampling rate.

[1] - O. Golokolenko and G. Schuller, "A FAST STEREO AUDIO SOURCE SEPARATION FOR MOVING SOURCES," in ASILOMAR, 2019

[2] - Robert Aichner, Herbert Buchner, Fei Yan, and Walter Kellermann, “A real-time blind source separation scheme and its application to reverberant and noisy acoustic environments,” Signal Processing, vol. 86, no.6, pp. 1260 – 1277, 2006, Applied Speech and Audio Processing.

Oleg Golokolenko, [email protected], August 2020.

vipchengrui / comparison-of-blind-source-separation-techniques Goto Github PK