Interspeech'2005 - Eurospeech

Lisbon, Portugal
September 4-8, 2005

The 2004 BBN/LIMSI 20xRT English Conversational Telephone Speech Recognition System

Rohit Prasad (1), Spyros Matsoukas (1), C.-L. Kao (1), Jeff Z. Ma (1), D.-X. Xu (1), T. Colthurst (1), O. Kimball (1), Richard Schwartz (1), Jean-Luc Gauvain (2), Lori Lamel (2), Holger Schwenk (2), G. Adda (2), F. Lefevre (2)

(1) BBN Technologies, Cambridge, MA, USA; (2) LIMSI-CNRS, Orsay, France

In this paper we describe the English Conversational Telephone Speech (CTS) recognition system jointly developed by BBN and LIMSI under the DARPA EARS program for the 2004 evaluation conducted by NIST. The 2004 BBN/LIMSI system achieved a word error rate (WER) of 13.5% at 18.3xRT (real-time as measured on Pentium 4 Xeon 3.4 GHz Processor) on the EARS progress test set. This translates into a 22.8% relative improvement in WER over the 2003 BBN/LIMSI EARS evaluation system, which was run without any time constraints. In addition to reporting on the system architecture and the evaluation results, we also highlight the significant improvements made at both sites.

Full Paper

Bibliographic reference.  Prasad, Rohit / Matsoukas, Spyros / Kao, C.-L. / Ma, Jeff Z. / Xu, D.-X. / Colthurst, T. / Kimball, O. / Schwartz, Richard / Gauvain, Jean-Luc / Lamel, Lori / Schwenk, Holger / Adda, G. / Lefevre, F. (2005): "The 2004 BBN/LIMSI 20xRT English conversational telephone speech recognition system", In INTERSPEECH-2005, 1645-1648.