Comparing Two Methods Based on Time-Frequency Analysis to Estimate the Instantaneous Mixing Matrix in Blind Audio Source Separation

Llerena C, Gil R, Álvarez L, Cuadra L, Ayllón D. Comparing Two Methods Based on Time-Frequency Analysis to Estimate the Instantaneous Mixing Matrix in Blind Audio Source Separation. In: AES Conference: 45th International Conference: Applications of Time-Frequency Processing in Audio. Audio Engineering Society; 2012. Paper 3-9. Available from: https://aes.org/publications/elibrary-page/?id=16206

@inproceedings{Llerena2012_16206,
  author = {Llerena, Cosme and Gil, Roberto and Álvarez, Lorena and Cuadra, Lucas and Ayllón, David},
  title = {{Comparing Two Methods Based on Time-Frequency Analysis to Estimate the Instantaneous Mixing Matrix in Blind Audio Source Separation}},
  booktitle = {AES Conference: 45th International Conference: Applications of Time-Frequency Processing in Audio},
  note = {Paper 3-9},
  year = {2012},
  month = mar,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=16206}
}

TY  - CPAPER
TI  - Comparing Two Methods Based on Time-Frequency Analysis to Estimate the Instantaneous Mixing Matrix in Blind Audio Source Separation
AU  - Llerena, Cosme
AU  - Gil, Roberto
AU  - Álvarez, Lorena
AU  - Cuadra, Lucas
AU  - Ayllón, David
T2  - AES Conference: 45th International Conference: Applications of Time-Frequency Processing in Audio
M1  - Paper 3-9
PY  - 2012
DA  - 2012/03/06
UR  - https://aes.org/publications/elibrary-page/?id=16206
PB  - Audio Engineering Society
LA  - en
AB  - This paper explores the feasibility of using two methods, based on time-frequency analysis, to estimate the instantaneous mixing matrix in Blind Audio Source Separation applications. These methods are: 1) the LOST line orientation separation technique algorithm; and 2) a method based on histograms. They exploit sparsity and independency of sources, and what is of key importance, work properly for under-determined cases. The comparative study shows that only when a) the number of mixtures is small, and b) the number of mixture samples is low, the histogram-based approach works better than the LOST algorithm. In the remaining problems, the LOST algorithm performs better than the histogram-based one, and is especially useful when the system has sucient number of samples to obtain an estimate of the mixing matrix.
ER  -