AUTOMATIC LEXICAL PRONUNCIATIONS GENERATION AND UPDATE

Most automatic speech recognizersuse a dictionary that maps words to one or more canonical pronunciations. Such en- tries are typically hand-written by lexical experts. In this research, we investigate a new approach for automatically generating lexical pronunciations using a linguistically mo- tivated subword model, and refining the pronunciations with spoken examples. The approach is evaluated on an isolated word recognition task with a 2k lexicon of restaurant and street names. A letter-to-sound model is first used to gen- erate seed baseforms for the lexicon. Then spoken utterances of words in the lexicon are presented to a subword recognizer and the top hypotheses are used to update the lexical base- forms. The spelling of each word is also used to constrain the subword search space and generate spelling-constrained baseforms. The results obtained are quite encouraging and indicate that our approach can be successfully used to learn valid pronunciations of new words.