Application of Super-paramagnetic Clustering in Speaker Clustering

The aim of speaker clustering is to partition speech segments into groups based on their similarities which have an important role in speech processing. There are many methods in speaker clustering. In this paper, we use super-paramagnetic clustering (SPC) algorithm for speaker clustering. The SPC algorithm makes no explicit assumptions about the structure of the data, and under very general and natural assumptions solves the clustering problem by evaluating thermal properties of a disordered magnet. The evaluation is conducted using the telephone speech segments. The experimental results show SPC algorithm is very effective in speaker clustering tasks