Robust speaker identification algorithms and results in noisy environments

Abstract

In this research, we have developed a robust speaker identification system, which involves mask estimation, gammatone features with bounded marginalization to deal with unreliable features, and Gaussian mixture model (GMM) for speaker identification. Extensive experiments using actual and synthesized conversations clearly demonstrated the performance of our algorithms under noisy conditions.