Unsupervised intra-speaker variability compensation based on Gestalt and model adaptation in speaker verification with telephone

In this paper an unsupervised compensation method based on Gestalt, ISVC, is proposed to address the problem of limited enrolling data and noise robustness in text-dependent speaker verification (SV). Reductions in EER and in the integral below the ROC curve as high as 20% or 40% and 30% or 60%, respectively, can be achieved by ISVC independently of the number of enrolling utterances. In contrast to model adaptation methods, ISVC is memoryless with respect to previous verification attempts. As shown here, unsupervised model adaptation can lead to substantial improvements in EER but is highly dependent on the sequence of client/impostor verification events. In adverse scenarios, such as massive impostor attacks and verification from alternated telephone line, unsupervised model adaptation might even provide reductions in verification accuracy when compared with the baseline system. In those cases, ISVC can even outperform adaptation schemes. It is worth emphasizing that ISVC and unsupe...

• Type: Journal
• Year: 2008
• Where: SPEECH
• Authors: Néstor Becerra Yoma, Claudio Garretón, Carlos Molina, Fernando Huenupán
• Keywords: SPEECH 2008, Text-dependent Speaker Verification, Unsupervised Compensation Method, Unsupervised Model Adaptation