View article

This paper presents the performance evaluation of i-vector and PLDA based speaker recognition system which incorporate quality measures function (QMF) in linear calibration. Evaluated on the recent NIST SRE’12 corpus, the linear calibration with QMF as the additional term shows a positive gain in the system performance compared to the conventional linear calibration with only two terms. Based on the equal error rate values measured from our I4U evaluation trial set, the QMF calibration approach shows 10–37% relative improvement compared to the conventional linear calibration. It is shown that by adding 1–2 extra parameters in the linear calibration through QMF approach, there is a potential to improve the calibration and discrimination performances of a speaker recognition system.