In open-set speaker identification systems a known phenomenon is that the false alarm (accept) error rate increases dramatically when increasing the number of registered speakers (models). In this paper, we demonstrate this phenomenon and suggest a solution using a new model-dependent score-normalization technique, called Top-norm. The Top-norm method was specifically developed to improve results of open-set speaker identification systems. Also, we suggest a score-normalization parameter adaptation technique. Experiments performed using speaker recognition corpora arc described and demonstrate that the new method outperforms other normalization methods.
|State||Published - 1 Dec 2006|
|Event||IEEE Odyssey 2006: Workshop on Speaker and Language Recognition - San Juan, Puerto Rico|
Duration: 28 Jun 2006 → 30 Jun 2006
|Conference||IEEE Odyssey 2006: Workshop on Speaker and Language Recognition|
|Period||28/06/06 → 30/06/06|