Word Alignment in Digital Talking Books Using WFSTs
This paper describes the motivation and the method that we used for aligning digital spoken books, and the results obtained both at a word level and at a phone level. This alignment will allow specific access interfaces for persons with special needs, and also tools for easily detecting and indexing units (words, sentences, topics) in the spoken books. The tool was implemented in a Weighted Finite State Transducer framework, which provides an efficient way to combine different types of knowledge sources, such as alternative pronunciation rules. With this tool, a 2-hour long spoken book was aligned in a single step in much less than real time.
KeywordsAcoustic Model Word Level Word Error Rate Terminal Symbol Word Alignment
Unable to display preview. Download preview PDF.
- 1.ANSI/NISO Z39.86— 2002 Specifications for the Digital Talking Book, http://www.niso.org/standards/index.html
- 3.Neto, J., Martins, C. and Almeida, L., A Large Vocabulary Continuous Speech Recognition Hybrid System for the Portuguese Language, in Proc. ICSLP 98, Sydney, Australia, 1998.Google Scholar
- 4.H. Meinedo and J. Neto, “Combination of acoustic models in continuous speech recognition hybridsystems”, In Proc. ICSLP 2000, Beijing, China, 2000.Google Scholar
- 5.H. Hermansky, N. Morgan, A. Baya and P. Kohn, “RASTA-PLP Speech Analysis Technique”, In Proc. ICASSP 92, San Francisco, USA, 1992.Google Scholar
- 7.M. Mohri, M. Riley, D. Hindle, A. Ljolje, F. Pereira, “Full Expansion of Context-Dependent Networks in Large Vocabulary Speech Recognition”, In Proc. ICASSP 98, Seattle, Washington, 1998.Google Scholar
- 8.C. Ribeiro, I. Trancoso and M. Viana, EUROM.1 Portuguese Database, Report of ESPRIT Project 6819 SAM-A, 1993.Google Scholar