University of the Basque Country (EHU) Systems for the NIST 2015 Language Recognition Evaluation

This paper briefly describes the language recognition systems developed by the Software Technology Working Group (http://gtts.ehu.es) of the University of the Basque Country (EHU) for the NIST 2015 Language Recognition Evaluation. The submitted systems follow the Total Variability Factor Analysis (iVector) approach. The systems use either Mel-Frequency Cepstral Coefficients (MFCC) or Phone Log-Likelihood Ratio (PLLR) as features [1]. Different modelings and scorings were applied on top of the ivectors, and several combinations of them were fused for the final submissions.