Prosodic Modeling in Large Vocabulary Mandarin Speech Recognition

The issue of incorporating prosodic information into speech recognition processes has emerged in recent years. In this work we present a complete framework for Mandarin speech recognition with prosodic modeling considering two-level hierarchical prosodic information for Mandarin Chinese. We developed a GMM-based, a decision-tree-based, and a hybrid approach. The best improvements in character recognition accuracy were obtained by the decision-tree-based prosodic models. This approach does NOT require a training corpus labeled with prosodic features, and works reasonably for a large-scale multi-speaker task.