Confusability measure based lexicon optimization for fast LVCSR decoding

Abstract

In this paper, we propose a lexicon optimization method based on confusability measure (CM) in order to reduce the decoding time for a large vocabulary continuous speech recognition (LVCSR) system. When lexicon is built or expanded for unseen words by using grapheme-to-phoneme (G2P) conversion, the lexicon size increases since G2P is generally realized by 1-to-N-best mapping. Thus, the proposed method prunes the confusable words in the lexicon by a CM that is defined a linguistic distance between two phonemic sequences. It is demonstrated from LVCSR experiments that the proposed lexicon optimization method achieves a relative real-time factor reduction of 23.13% on a task on the Wall Street Journal, compared to the 1-to-4-best G2P converted lexicon approach.