The IBM 2008 GALE Arabic speech transcription system

Author

Saon, George ; Soltau, Hagen ; Chaudhari, Upendra ; Chu, Stephen ; Kingsbury, Brian ; Kuo, Hong-Kwang ; Mangu, Lidia ; Povey, Daniel

Author_Institution

IBM T. J. Watson Res. Center, Yorktown Heights, NY, USA

fYear

2010

fDate

14-19 March 2010

Firstpage

4378

Lastpage

4381

Abstract

This paper describes the Arabic broadcast transcription system fielded by IBM in the GALE Phase 3.5 machine translation evaluation. Key advances compared to our Phase 2.5 system include improved discriminative training, the use of Subspace Gaussian Mixture Models (SGMM), neural network acoustic features, variable frame rate decoding, training data partitioning experiments, unpruned n-gram language models and neural network language models. These advances were instrumental in achieving a word error rate of 8.9% on the evaluation test set.

Keywords

Gaussian processes; acoustic signal processing; decoding; error statistics; language translation; natural language processing; neural nets; speech coding; speech recognition; Arabic broadcast transcription system; GALE Phase 3.5 machine translation evaluation; IBM 2008 GALE Arabic speech transcription system; n-gram language model; neural network acoustic features; neural network language model; speech recognition; subspace Gaussian mixture model; training data partitioning; variable frame rate decoding; word error rate; Acoustic testing; Broadcasting; Decoding; Loudspeakers; Maximum likelihood linear regression; Natural languages; Neural networks; Phase estimation; Speech analysis; Speech recognition; Speech recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics Speech and Signal Processing (ICASSP), 2010 IEEE International Conference on

Conference_Location

Dallas, TX

ISSN

1520-6149

Print_ISBN

978-1-4244-4295-9

Electronic_ISBN

1520-6149

Type

conf

DOI

10.1109/ICASSP.2010.5495640

Filename

5495640