Text-independent Speaker Recognition by Speaker-specific GMM and Speaker Adapted Syllable-based HMM

摘要

We present a new text-independent speaker recognition method by combining speaker-specific Gaussian Mixture Model(GMM) with syllable-based HMM adapted by MLLR or MAP. The robustness of this speaker recognition method for speaking style's change was evaluated. The speaker identification experiment using NTT database which consists of sentences data uttered at three speed modes (normal, fast and slow) by 35 Japanese speakers(22 males and 13 females) on five sessions over ten months was conducted. Each speaker uttered only 5 training utterances. We obtained the accuracy of 100% for text-independent speaker identification. This result was superior to some conventional methods for the same database.