Independent Low-Rank Matrix Analysis Based on Parametric Majorization-Equalization Algorithm
This is an incremental improvement for blind source separation in audio or signal processing.
The paper tackles the problem of poor local solutions in independent low-rank matrix analysis (ILRMA) due to differing convergence speeds between spatial and source models, by introducing a parameter to control the source model's convergence speed and finding an optimal balance, resulting in improved performance.
In this paper, we propose a new optimization method for independent low-rank matrix analysis (ILRMA) based on a parametric majorization-equalization algorithm. ILRMA is an efficient blind source separation technique that simultaneously estimates a spatial demixing matrix (spatial model) and the power spectrograms of each estimated source (source model). In ILRMA, since both models are alternately optimized by iterative update rules, the difference in the convergence speeds between these models often results in a poor local solution. To solve this problem, we introduce a new parameter that controls the convergence speed of the source model and find the best balance between the optimizations in the spatial and source models for ILRMA.