Study on Cross-Lingual Adaptation of a Czech LVCSR System towards Slovak
This paper deals with cross-lingual adaptation of a Large Vocabulary Continuous Speech Recognition (LVCSR) system between two similar Slavic languages – from Czech to Slovak. The proposed adaptation scheme is performed in two consecutive phases and it is focused on acoustic modeling and phoneme and pronunciation mapping. It also utilizes language similarities between the source and the target language and speaker adaptation approaches. Presented experimental results show that the proposed cross-lingual adaptation approach yields to reduction of Word Error Rate (WER) from 12.8 % to 8.1 % in the voice dictation task.
Keywordsspeech recognition cross-lingual adaptation speaker adaptation Slavic languages
Unable to display preview. Download preview PDF.
- 1.Nouza, J., Zdansky, J., Cerva, P., Kolorenc, J.: Continual On-line Monitoring of Czech Spoken Broadcast Programs. In: Proceedings of International Conference on Spoken Language Processing (ICSLP) 2006, Pittsburgh, USA, pp. 1650–1653 (September 2006)Google Scholar
- 2.Bayeh, R., Lin, S., Chollet, G., Mokbel, C.: Towards multilingual speech recognition using data driven source/target acoustical units association. In: Proceedings of IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2004, Montreal, Quebec, Canada, pp. 521–524 (2004)Google Scholar
- 3.Lin, H., Deng, L., Yu, D., Gong, Y., Acero, A., Lee, C.H.: A study on multilingual acoustic modeling for large vocabulary ASR. In: Proceedings of IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2009, Taipei, Taiwan, pp. 4333–4336 (2009)Google Scholar
- 4.EU Web Pages, http://eur-lex.europa.eu/en/treaties/index.htm
- 5.Kral, A., Sabol, J.: Phonetics and Phonology. SPN, Bratislava (1989) (in Slovak)Google Scholar
- 6.Ivanecky, J.: Automatic speech transcription and segmentation. PhD thesis, Košice (2003) (in Slovak)Google Scholar