Progress in transcription of broadcast News using Byblos

Authors:
Long Nguyen;Spyros Matsoukas;Jason Davenport;Francis Kubala;Richard Schwartz;John Makhoul
Affiliations:
BBN Technologies, Verizon Communications Inc., 70 Fawcett Street, Cambridge, MA;BBN Technologies, Verizon Communications Inc., 70 Fawcett Street, Cambridge, MA;BBN Technologies, Verizon Communications Inc., 70 Fawcett Street, Cambridge, MA;BBN Technologies, Verizon Communications Inc., 70 Fawcett Street, Cambridge, MA;BBN Technologies, Verizon Communications Inc., 70 Fawcett Street, Cambridge, MA;BBN Technologies, Verizon Communications Inc., 70 Fawcett Street, Cambridge, MA
Venue:
Speech Communication
Year:
2002

Citing 6
Cited 2

Broadcast News Transcription

ICASSP '97 Proceedings of the 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP '97) -Volume 1 - Volume 1
The harpy speech recognition system.

The harpy speech recognition system.
Search algorithms for software-only real-time recognition with very large vocabularies

HLT '93 Proceedings of the workshop on Human Language Technology
Recent experiments in large vocabulary conversational speech recognition

ICASSP '99 Proceedings of the Acoustics, Speech, and Signal Processing, 1999. on 1999 IEEE International Conference - Volume 01
Single-tree method for grammar-directed search

ICASSP '99 Proceedings of the Acoustics, Speech, and Signal Processing, 1999. on 1999 IEEE International Conference - Volume 02
Towards a robust real-time decoder

ICASSP '99 Proceedings of the Acoustics, Speech, and Signal Processing, 1999. on 1999 IEEE International Conference - Volume 02

Corrigendum to "Editorial"

Speech Communication
Structuring broadcast audio for information access

EURASIP Journal on Applied Signal Processing

Quantified Score

Hi-index	0.00

Visualization

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.