Time-frequency clustering with weighted and contextual information for convolutive blind source separation

In this paper we investigate the use of observation weights and contextual time-frequency information for clustering-based blind source separation. Previous clustering-based approaches have successfully used clustering techniques to estimate time-frequency separation masks; however, these approaches generally disregard the structured nature of speech signals. Motivated by the homogenous behavior of speech signals, we propose to modify the established fuzzy c-means algorithm to bias the clustering results in favor of cluster membership homogeneity within localized neighborhoods in the time-frequency space. This problem can be solved by using a two-stage algorithm: firstly, the estimation of data weights to indicate the reliability of each data point, and secondly, the integration of local contextual information into the cluster update equations from neighboring time-frequency slots. The proposed algorithm is evaluated in a three-fold manner using simulated, real recordings and public benchmark data; notable improvement in source separation performance over previous clustering approaches was achieved.