Croatian Large Vocabulary Automatic Speech Recognition (CROSBI ID 173928)
Prilog u časopisu | izvorni znanstveni rad | međunarodna recenzija
Podaci o odgovornosti
Martinčić-Ipšić, Sanda ; Pobar, Miran ; Ipšić, Ivo
engleski
Croatian Large Vocabulary Automatic Speech Recognition
This paper presents procedures used for development of a Croatian large vocabulary automatic speech recognition system. The proposed acoustic model is based on context-dependent triphone hidden Markov models and Croatian phonetic rules. Different acoustic and language models, developed using a large collection of Croatian speech, are discussed and compared. The paper proposes the best feature vectors and acoustic modeling procedures for Croatian automatic speech recognition where the lowest word error rates were achieved. In addition, Croatian language modeling procedures are evaluated and adopted for speaker independent speech recognition. Presented experiments and results show that the proposed approach for automatic speech recognition using context-dependent acoustic modeling based on Croatian phonetic rules and a parameter tying procedure can be used for efficient Croatian large vocabulary speech recognition with word error rates below 5%.
acoustic modeling; automatic speech recognition; context-dependent acoustic units; language modeling
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o izdanju
52 (2)
2011.
147-157
objavljeno
0005-1144
Povezanost rada
Računarstvo, Informacijske i komunikacijske znanosti