This paper presents an approach with ensemble classifiers using unsupervised data selection for speaker recognition. Ensemble learning is a type of machine learning that applies a combination of several weak learners to achieve accurate and improved performance than a single learner. Based on its acoustic characteristics, the speech utterance is divided into several subsets using unsupervised data selection methods. The ensemble classifiers are then trained with these non-overlapping subsets of speech data to improve the recognition accuracy. Our experiments on the 2008 and 2010 NIST Speaker Recognition Evaluation datasets show that using ensemble classifiers substantially reduces DCF.

Index Terms: speaker recognition, ensemble classifier, unsupervised data selection