16th Annual Conference of the International Speech Communication Association

Dresden, Germany
September 6-10, 2015

Blind Score Normalization Method for PLDA Based Speaker Recognition

Danila Doroshin, Nikolay Lubimov, Marina Nastasenko, Mikhail Kotov

Stel Computer Systems, Russia

Probabilistic Linear Discriminant Analysis (PLDA) has become state-of-the-art method for modeling i-vector space in speaker recognition task. However the performance degradation is observed if enrollment data size differs from one speaker to another. This paper presents a solution to such problem by introducing new PLDA scoring normalization technique. Normalization parameters are derived in a blind way, so that, unlike traditional
    ZT-norm, no extra development data is required. Moreover, proposed method has shown to be optimal in terms of detection cost function. The experiments conducted on NIST SRE 2014 database demonstrate an improved accuracy in a mixed enrollment number condition.

Full Paper

Bibliographic reference.  Doroshin, Danila / Lubimov, Nikolay / Nastasenko, Marina / Kotov, Mikhail (2015): "Blind score normalization method for PLDA based speaker recognition", In INTERSPEECH-2015, 210-213.