HMM-BASED TEXT-TO-PHONEME PARSER AND METHOD FOR TRAINING SAME

摘要

AN IIMM-BASED TEXT-TO-PHONEME PARSER (24) USES PROBABILITY INFORMATION WITHIN A PROBABILITY DATABASE (26) TO GENERATE ONE OR MORE PHONEME STRINGS FOR A WRITTEN INPUT WORD. TECHNIQUES FOR TRAINING THE TEXT-TO-PHONEME PARSER (24) ARE PROVIDED. FIG. 2