Noise and Channel Distortion Robust ASR System for DARPA SPINE2 Task

Konstantin MARKOV; Tomoko MATSUI; Rainer GRUHN; Jinsong ZHANG; Satoshi NAKAMURA

Noise and Channel Distortion Robust ASR System for DARPA SPINE2 Task

Konstantin MARKOV, Tomoko MATSUI, Rainer GRUHN, Jinsong ZHANG, Satoshi NAKAMURA

Full Text Views

0

Share
Cite this

Summary :

This paper presents the ATR speech recognition system designed for the DARPA SPINE2 evaluation task. The system is capable of dealing with speech from highly variable, real-world noisy conditions and communication channels. A number of robust techniques are implemented, such as differential spectrum mel-scale cepstrum features, on-line MLLR adaptation, and word-level hypothesis combination, which led to a significant reduction in the word error rate.

Publication: IEICE TRANSACTIONS on Information Vol.E86-D No.3 pp.497-504

Publication Date: 2003/03/01

Publicized

Online ISSN

DOI

Type of Manuscript: Special Section PAPER (Special Issue on Speech Information Processing)

Category: Robust Speech Recognition and Enhancement

Cite this

Copy

Konstantin MARKOV, Tomoko MATSUI, Rainer GRUHN, Jinsong ZHANG, Satoshi NAKAMURA, "Noise and Channel Distortion Robust ASR System for DARPA SPINE2 Task" in IEICE TRANSACTIONS on Information, vol. E86-D, no. 3, pp. 497-504, March 2003, doi: .
Abstract: This paper presents the ATR speech recognition system designed for the DARPA SPINE2 evaluation task. The system is capable of dealing with speech from highly variable, real-world noisy conditions and communication channels. A number of robust techniques are implemented, such as differential spectrum mel-scale cepstrum features, on-line MLLR adaptation, and word-level hypothesis combination, which led to a significant reduction in the word error rate.
URL: https://globals.ieice.org/en_transactions/information/10.1587/e86-d_3_497/_p

Copy

@ARTICLE{e86-d_3_497,
author={Konstantin MARKOV, Tomoko MATSUI, Rainer GRUHN, Jinsong ZHANG, Satoshi NAKAMURA, },
journal={IEICE TRANSACTIONS on Information},
title={Noise and Channel Distortion Robust ASR System for DARPA SPINE2 Task},
year={2003},
volume={E86-D},
number={3},
pages={497-504},
abstract={This paper presents the ATR speech recognition system designed for the DARPA SPINE2 evaluation task. The system is capable of dealing with speech from highly variable, real-world noisy conditions and communication channels. A number of robust techniques are implemented, such as differential spectrum mel-scale cepstrum features, on-line MLLR adaptation, and word-level hypothesis combination, which led to a significant reduction in the word error rate.},
keywords={},
doi={},
ISSN={},
month={March},}

Copy

TY - JOUR
TI - Noise and Channel Distortion Robust ASR System for DARPA SPINE2 Task
T2 - IEICE TRANSACTIONS on Information
SP - 497
EP - 504
AU - Konstantin MARKOV
AU - Tomoko MATSUI
AU - Rainer GRUHN
AU - Jinsong ZHANG
AU - Satoshi NAKAMURA
PY - 2003
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E86-D
IS - 3
JA - IEICE TRANSACTIONS on Information
Y1 - March 2003
AB - This paper presents the ATR speech recognition system designed for the DARPA SPINE2 evaluation task. The system is capable of dealing with speech from highly variable, real-world noisy conditions and communication channels. A number of robust techniques are implemented, such as differential spectrum mel-scale cepstrum features, on-line MLLR adaptation, and word-level hypothesis combination, which led to a significant reduction in the word error rate.
ER -