The phoneme set influence for Lithuanian speech commands recognition accuracy

Abstract

The phoneme set influence for Lithuanian speech commands recognition accuracy is investigated. Four phoneme sets are discussed. LIEPA speech corpus for training of Acoustic Model is used. The phonetic representation of corpus transcriptions is generated by grapheme-to-phoneme transformation rules. Rule based transformations for Lithuanian language is proposed. Recognition engine with CMU Pocketsphinx decoder is used. Experimental investigation with fixed size test corpus demonstrates that 36 (FZ1.3) phonemes set shows the bests result.