1. Progress in transcription of Broadcast News using Byblos
- Author
-
Spyros Matsoukas, J. Makhoul, Richard Schwartz, Francis Kubala, Jason Davenport, and Long Nguyen
- Subjects
Linguistics and Language ,Computer science ,Communication ,Speech recognition ,Speaker recognition ,Language and Linguistics ,Computer Science Applications ,Search algorithm ,Modeling and Simulation ,Computer Vision and Pattern Recognition ,Transcription (software) ,Hidden Markov model ,Software ,Speaker adaptation - Abstract
In this paper, we describe our progress during the last four years (1995-1999) in automatic transcription of broadcast news from radio and television using the BBN Byblos speech recognition system. Overall, we achieved steady progress as reflected through the results of the last four DARPA Hub-4 evaluations, with word error rates of 42.7%, 31.8%, 20.4% and 14.7% in 1995, 1996, 1997 and 1998, respectively. This progress can be attributed to improvements in acoustic modeling, channel and speaker adaptation, and search algorithms, as well as dealing with specific characteristics of the real-life variable speech found in broadcast news. Besides improving recognition accuracy, we also succeeded in developing several algorithms to achieve close-to-real-time recognition speed without a significant sacrifice in recognition accuracy.
- Published
- 2002
- Full Text
- View/download PDF