Data sampling approaches for GMM supervector based speaker verification GKM süpervektör tabanli konuşmaci doǧrulamada veri̇ örnekleme yaklaşimlari
2011 IEEE 19th Signal Processing and Communications Applications Conference, SIU 2011, Antalya, Türkiye, 20 - 22 Nisan 2011, ss.562-565, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/siu.2011.5929716
- Basıldığı Şehir: Antalya
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.562-565
- Boğaziçi Üniversitesi Adresli: Evet
Özet
GMM supervectors are among the most popular feature sets used in SVM-based text-independent speaker verification. Most of the studies represent speaker characteristics obtained from a long recording with a single supervector in the SVM space. Working on the NIST SRE'10 dataset, this study compares the effect of two sampling methods to increase the number of supervectors, on the verification performance. Dominance of positive and negative classes on model construction is investigated. © 2011 IEEE.