A Fast, Sequential Decoding Algorithm with Application to Speaker Verification

In [1], we proposed a fast, sequential decoding algorithm for left-to-right HMM. The algorithm is based on a sequential detection scheme which is asymptotically optimal in the sense of detecting a possible change in distribution as reliably and quickly as possible. As a sequential scheme, the algorithm can determine a state boundary in a few time steps after it occurs. In this paper, the algorithm is evaluated in a fixed-phrase speaker verification system on a database with 23,578 utterances recorded from 100 speakers. The experimental results show that the accuracy of the proposed algorithm is close to the Viterbi algorithm while the speed is about 7.6 times faster than the Viterbi algorithm on speaker-dependent model decoding. The results indicate that the proposed algorithm can be applied to speaker verification, speaker identification, utterance verification, audio segmentation, voice/silence detection and many other applications.