Blind Audio Source Separation (BASS) is a crucial problem in the field of speech and audio processing. Its goal is to separate different sources from their mixture. In the case of audio mining, trying to analyze the content of the audio signal generally consists of designing a print or pattern of a given sound to recognize. Thus, one has to admit that when the signal is a mixture, it becomes difficult to extract suitable features allowing the design of a particular sound.

Let us denote a mixture of source signals received by sensors .

The mixture in practice is convolutive and can be expressed by the following equation:

.

where is the -th mixed signal, are the coefficients of the mixing filter and an additive noise of the -th sensor. The -transform of the above equation is:

.

The separation aims at finding an estimate of the original signal . This is performed by estimating the mixing filter inverse via . Therefore the -transform formulation of the separation model is:

,

where and .

Different algorithms can be used to perform BASS; these include:

? Principal Component Analysis (PCA)

? Independent Component Analysis (ICA) [1]

? Non-negative Matrix Factorization (NMF) [2].

[1] P. Comon “Independent Component Analysis: a new concept?” 1994.

[2] B. Wang, M. D. Plumbey. Musical audio stream separation by N-negative Matrix Factorization. EUSIPCO 2005.

You may also like