Speaker indexing based on speaker model selection and automatic speech recognition in discussions

摘要

This paper addresses unsupervised speaker indexing for discussion audio archives. In discussions, the speaker changes frequently, thus the duration of utterances is very short and its variation is large, which causes significant problems in applying conventional methods such as model adaptation and Variance-BIC (Bayesian Information Criterion) methods. We propose a flexible framework that selects an optimal speaker model (GMM or VQ) based on the BIC according to the duration of utterances. When the speech segment is short, the simple and robust VQ-based method is expected to be chosen, while GMM will be reliably trained for long segments. For a discussion archive, it is demonstrated that the proposed method achieves higher indexing performance than that of conventional methods. The speaker index is useful for speaker adaptation of the acoustic model, which improves