An EM Algorithm for Training Wideband Acoustic Models from Mixed-Bandwidth Training Data

One serious difficulty in the deployment of wideband speech recognition systems for new tasks is the expense in both time and cost of obtaining sufficient training data. A more economical approach is to collect telephone speech and then restrict the application to operate at the telephone bandwidth. However, this generally results in suboptimal performance compared to a wideband recognition system. In this paper, we propose a novel EM algorithm in which wideband acoustic models are trained using a small amount of wideband speech augmented by a larger amount of narrowband speech. Experiments performed using wideband speech and telephone speech demonstrate that the proposed mixedbandwidth training algorithm results in significant improvements in recognition accuracy over conventional training strategies when the amount of wideband data is limited.