An experimental study of speaker verification sensitivity to computer voice-altered imposters

B.L. Pellom, J.H.L. Hansen
1999 1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258)  
This paper investigates the relative sensitivity o f a G M M based voice veri cation algorithm to computer voice altered imposters. First, a new trainable speech s y n thesis algorithm based on trajectory models of the speech Line Spectral Frequency LSF parameters is presented in order to model the spectral characteristics of a target voice. A GMM based speaker veri er is then constructed for the 138 speaker YOHO database and shown to have an initial equal error rate EER of 1.45 for the case of
more » ... .45 for the case of casual imposter attempts using a single combination lock phrase test. Next, imposter voices are automatically altered using the synthesis algorithm to mimic the customer's voice. After voice transformation, the false acceptance rate is shown to increase from 1.45 to over 86 if the baseline EER threshold is left unmodi ed. Furthermore, at a customer false rejection rate of 25, the false acceptance rate for the voice altered imposter remains as high as 34.6.
doi:10.1109/icassp.1999.759801 dblp:conf/icassp/PellomH99 fatcat:bdhnoa6ksffd7gbtlew22ffr7e