Recent experiments in large vocabulary conversational speech recognition

Authors:
J. Billa;T. Colhurst;A. El-Jaroudi;R. Iyer;K. Ma;S. Marsoukas;C. Quillen;F. Richardson;M. Siu;G. Zavaliagkos;H. Gish
Affiliations:
BBN Technol., Cambridge, MA, USA;-;-;-;-;-;-;-;-;-;-
Venue:
ICASSP '99 Proceedings of the Acoustics, Speech, and Signal Processing, 1999. on 1999 IEEE International Conference - Volume 01
Year:
1999

Citing 0
Cited 3

Progress in transcription of broadcast News using Byblos

Speech Communication
Developing HMM-Based Recognizers with ESMERALDA

TSD '99 Proceedings of the Second International Workshop on Text, Speech and Dialogue
Reduced feature-set based parallel CHMM speech recognition systems

Information Sciences—Informatics and Computer Science: An International Journal - Special issue: Spoken language analysis, modeling and recognition-statistical and adaptive connectionist approaches

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper describes the improvements that resulted in the 1998 Byblos large vocabulary conversational speech recognition (LVCSR) system. Salient among these improvements are: improved signal processing, improved hidden Markov model (HMM) topology, use of quinphone context, introduction of diagonal speaker adapted training (DSAT), incorporation of variance adaptation in the MLLR framework, improvements in language modeling, increase in lexicon size and combination of multiple systems. These changes resulted in about a 7% absolute reduction in word error rates on a balanced Switchboard/Callhome English test set.