Automatic Phonetic Baseform Determination

Automatic Phonetic Baseform Determination L. R. Bahl, S. Das, P. V. deSouza, M. Epstein, R. L. Mercer, B. Merialdo, D. Nahamoo, M. A. Picheny, J. Powell Continuous Speech Recognition Group, Computer Sciences Department IBM Research Division, Thomas J. Watson Research Center P. O. Box 704, Yorktown Heights, NY 10598 ABSTRACT Phonetic baseforms are the basic recognition units in most large vocabulary speech recognition systems. These base- forms are usually determined by hand once a vocabulary is chosen and not modified thereafter. However, many applica- tions of speech recognition, such as dictation transcription, are hampered by a fixed vocabulary and require the user be able to add new words to the vocabulary. At least one phonetic base- form must be assigned to each new word to properly integrate the word into the recognition system. Dictionary lookup is of- ten unsuccessful in determining a phonetic baseform because new words are often names or task-specific jargon; also, talk- ers tend to have idiosyncratic pronunciations for a substantial fraction of words. This paper describes a series of experiments in which the phonetic baseform is deduced automatically for new words by utilizing actual utterances of the new word in conjunction with a set of automatically derived spelling-to- sound rules. We evaluated recognition performance on new words spoken by two different talkers when the phonetic base- forms were extracted via the above approach. The error rates on these new words were found to be comparable to or better than when the phonetic baseforms were derived by hand, thus validating the basic approach. 1 Introduction Phonetic baseforms are the basic recognition units in most large vocabulary speech recognition systems. These base- forms are usually determined by hand once a vocabulary is chosen and not modified thereafter. However, many applica- tions of speech recognition, such as dictation transcription, are hampered by a fixed vocabulary and require the user be able to add new words to the vocabulary. At least one phonetic base- form must be assigned to each new word to properly integrate the word into the recognition system 1. This paper describes a seres of experiments in which the phonetic baseform is deduced automatically for a new word given its spelling by utilizing actual utterances of the new word in conjunction with a set of automatically derived spelling-to-sound rules. Most previous attempts to construct a phonetic baseform for a new word have been in the area of speech synthesis. Klatt t One does not necessarily require a phonetic baseform to add a new word to a recognition system [ I, 2], but if the system is odginaUy based on phonetic baseforms, the system structure is complicated by having more than one type of recognition model present. [3] reviews various approaches to this problem; most solu- tions involve some set of human-derived rules combined with a dictionary lookup procedure for exceptions. The best sys- tems seem to produce correct baseforms for 95%-98% of the words in running text. We believe that such systems are inad- equate to handle …

Content truncated to 500 words