Speaker-adapted training on the Switchboard Corpus

Speaker adaptation is the process of transforming some speaker-independent acoustic model in such a way as to more closely match the characteristics of a particular speaker. It has been shown by several researchers to be an e ective means of improving the performance of large vocabulary continuous speech recognition systems. Until very recently speaker adaptation has been used exclusively as a part of the recognition process. This is undesireable inasmuch as it leads to a mismatched condition between test and training, and hence sub-optimal recognition performance. Very recently, there has been a growing interest in applying speaker-adaptation techniques to HMM training in order to alleviate the training/test mismatch. In prior work, we presented an iterative scheme for determining the maximum likelihood solution for the set of speaker-independent means and variances when speaker-dependent adaptation is performed during HMM training. In the present work, we shall investigate speci c issues encountered in applying this general framework to the task of improving recognition performance on the Switchboard Corpus.