Title :
Speaker recognition with rival penalized EM training
Author :
Matza, Avi ; Bistritz, Yuval
Author_Institution :
Sch. of Electr. Eng., Tel-Aviv Univ., Tel-Aviv, Israel
Abstract :
The paper considers speaker recognition with Gaussian mixture models trained by a rival penalized EM (RPEM) algorithm. Although RPEM was applied successfully to several pattern recognition problems, our attempt to apply the algorithm in its original form to speaker recognition was not successful. We modified it by adding a discriminative threshold to prevent over penalty on mixture components, and using it with batches of feature vectors rather than the original incremental mode. We applied the modified RPEM to train speaker models with the number of Gaussian mixture components adapted individually to each speaker and used it to perform some basic speaker recognition experiments. The experiments are very reassuring about using the modified RPEM as a training method for GMM based speaker recognition. In settings with limited amount of training data, not only that the algorithm showed nice convergence to reduced order speaker models, but the resulting reduced models achieved better recognition rates than the initial higher order models.
Keywords :
Gaussian processes; pattern recognition; speaker recognition; Gaussian mixture component; Gaussian mixture model; incremental mode; pattern recognition problem; recognition rate; reduced order speaker model; rival penalized EM algorithm; rival penalized EM training; speaker recognition; Convergence; Data models; Mel frequency cepstral coefficient; Speaker recognition; Testing; Training; Vectors; GMM; RPEM; speaker recognition;
Conference_Titel :
Machine Learning for Signal Processing (MLSP), 2011 IEEE International Workshop on
Conference_Location :
Santander
Print_ISBN :
978-1-4577-1621-8
Electronic_ISBN :
1551-2541
DOI :
10.1109/MLSP.2011.6064597