Research on the HMM-UBM and Short Text Based Speaker Verification
Zhiqiang Yao · Information and Computation · 2004
In this paper, a simple and effective threshold- se tting method for text-dependent speaker verification is proposed. The Gaussian mixture model(GMM) is also used as the background model to simulate the speech o f unpredicted impostors and outperforms the conventional HMM(hidden Markov mode l) background model in the case of diversified pseudo text. Based on these two t echnologies, a remote text-dependent speaker verification system over Internet is proposed. This system is evaluated in both experimental and real condition an d shows satisfied verification performance.