EVALITA 2009 Speaker Identity Verication \Application" Track |Organizer's Report|
暂无分享,去创建一个
This is an analysis and comparison of the accuracy of the seven primary systems submitted to the EVALITA 2009 Speaker Iden- tity Verication \Application" Track. We analyze performance accord- ing to test-segment duration, mode of enrollment and telephone types. In particular, we highlight two important problems: the diculty with mismatched telephone types in enrollment and test; and the diculty of consistent score calibration across dierent conditions.
[1] Tom Fawcett,et al. Robust Classification for Imprecise Environments , 2000, Machine Learning.
[2] Alvin F. Martin,et al. NIST 2008 speaker recognition evaluation: performance across telephone and room microphone channels , 2009, INTERSPEECH.
[3] Alvin F. Martin,et al. The DET curve in assessment of detection task performance , 1997, EUROSPEECH.