A Speaker Clustering Adaptive Algorithm Based on Gaussian Mixture Model

In this paper,a speaker clustering adaptive algorithm based on the Gaussian Mixture Model is provided The Gaussian Mixture Model is described and how to get the GMM parameters is also introduced The speaker clustering adaptive algorithm is applied to the keyword spotting systems The experimental result shows that an average detection rate improvement of 3 2% can be obtained by using the proposed adaptive algorithm with small computation and fast implementation This algorithm can be expected to be applied to a larger group of people and shows a wide prospect of application in small to intermediate range vocabulary recognition systems