Towards Automatic Transcription of Spontaneous Czech Speech in the MALACH Project
Our paper discusses the progress achieved during one-year effort with building the Czech LVCSR system for an automatic transcription of spontaneously pronounced testimonies in the MALACH project . The difficulty of this task stems from the highly inflectional nature of the Czech language and is further multiplied by the presence of many colloquial words in spontaneous Czech speech and also by the need to handle emotional speech filled with disfluencies, heavy accents, age-related coarticulation and language switching. In this paper we concetrate mainly on the acoustic issues – the proper choice of the front-end parameterization, handling the non-speech events in acoustic modeling and especially the unsupervised usage of the MLLR adaptation technique. A method for selecting suitable language model data is also briefly mentioned.
KeywordsAutomatic Speech Recognition Acoustic Model Emotional Speech Speaker Adaptation Language Switching
Unable to display preview. Download preview PDF.
- 2.Psutka, J., Ircing, P., Psutka, J.V., Radová, V., Byrne, V., Hajič, J., Gustman, S., Ramabhadran, B.: Automatic Transcription of Czech Language Oral History in the MALACH Project: Resources and Initial Experiments. In: Proceedings of TSD 2002, Brno (2002)Google Scholar
- 3.Psutka, J., Ircing, P., Psutka, J.V., Radová, V., Byrne, W., Hajič, J., Mírovský, J., Gustman, S.: Large Vocabulary ASR for Spontaneous Czech in the MALACH Project. Submitted to Eurospeech (2003)Google Scholar
- 4.Psutka, J., Müller, L., Psutka, J.V.: Comparison of MFCC and PLP Parameterization in the Speaker Independent Continuous Speech Recognition Task. In: Proceedings of Eurospeech 2001, Aalborg (2001)Google Scholar
- 5.Psutka, J., Radová, V., Müller, L., Matoušek, J., Ircing, P., Graff, D.: Large Broadcast News and Read Speech Corpora of Spoken Czech. In: Proceedings of Eurospeech 2001, Aalborg (2001)Google Scholar
- 6.Young, S., et al.: The HTK Book. Entropic Inc., Cambridge (1999)Google Scholar
- 7.Stolcke, A.: SRILM - an Extensible Language Modeling Toolkit. In: Proceedings of ICSLP 2002, Denver (2002)Google Scholar
- 8.Mohri, M., Pereira, F., Riley, M.: Weighted Finite-State Transducers in Speech Recognition. In: Proceedings of ASR 2000, International Workshop on Automatic Speech Recognition: Challenges for the Next Millennium, Paris (2000)Google Scholar