ISCA Archive Interspeech 2004
ISCA Archive Interspeech 2004

SVM modeling of "SNERF-grams" for speaker recognition

Elizabeth Shriberg, Luciana Ferrer, Anand Venkataraman, Sachin Kajarekar

We describe a new approach to modeling idiosyncratic prosodic behavior for automatic speaker recognition. The approach computes prosodic features by syllable, and models syllable-feature sequences using support vector machines. We evaluate performance on development data for a system submitted to the NIST 2004 Speaker Recognition evaluation. Results show that SNERF-grams provide significant performance gains when combined with a state-of-the-art cepstral system as well as with both prosodic and word-based systems that model long-range features.