Speech segmentation and clustering problem based on an unknown-multiple N signal source model - an application to segmented speech clustering based on speaker features.