Simultaneous Adaptation of Echo Cancellation and Spectral Subtraction for In-Car Speech Recognition

Osamu ICHIKAWA; Masafumi NISHIMURA

doi:10.1093/ietfec/e88-a.7.1732

Simultaneous Adaptation of Echo Cancellation and Spectral Subtraction for In-Car Speech Recognition

Osamu ICHIKAWA, Masafumi NISHIMURA

Full Text Views

0

Share
Cite this

Summary :

Recently, automatic speech recognition in a car has practical uses for applications like car-navigation and hands-free telephone dialers. For noise robustness, the current successes are based on the assumption that there is only a stationary cruising noise. Therefore, the recognition rate is greatly reduced when there is music or news coming from a radio or a CD player in the car. Since reference signals are available from such in-vehicle units, there is great hope that echo cancellers can eliminate the echo component in the observed noisy signals. However, previous research reported that the performance of an echo canceller is degraded in very noisy conditions. This implies it is desirable to combine the processes of echo cancellation and noise reduction. In this paper, we propose a system that uses echo cancellation and spectral subtraction simultaneously. A stationary noise component for spectral subtraction is estimated through the adaptation of an echo canceller. In our experiments, this system significantly reduced the errors in automatic speech recognition compared with the conventional combination of echo cancellation and spectral subtraction.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E88-A No.7 pp.1732-1738

Publication Date: 2005/07/01

Publicized

Online ISSN

DOI: 10.1093/ietfec/e88-a.7.1732

Type of Manuscript: Special Section PAPER (Special Section on Multi-channel Acoustic Signal Processing)

Category: Speech Enhancement

Cite this

Copy

Osamu ICHIKAWA, Masafumi NISHIMURA, "Simultaneous Adaptation of Echo Cancellation and Spectral Subtraction for In-Car Speech Recognition" in IEICE TRANSACTIONS on Fundamentals, vol. E88-A, no. 7, pp. 1732-1738, July 2005, doi: 10.1093/ietfec/e88-a.7.1732.
Abstract: Recently, automatic speech recognition in a car has practical uses for applications like car-navigation and hands-free telephone dialers. For noise robustness, the current successes are based on the assumption that there is only a stationary cruising noise. Therefore, the recognition rate is greatly reduced when there is music or news coming from a radio or a CD player in the car. Since reference signals are available from such in-vehicle units, there is great hope that echo cancellers can eliminate the echo component in the observed noisy signals. However, previous research reported that the performance of an echo canceller is degraded in very noisy conditions. This implies it is desirable to combine the processes of echo cancellation and noise reduction. In this paper, we propose a system that uses echo cancellation and spectral subtraction simultaneously. A stationary noise component for spectral subtraction is estimated through the adaptation of an echo canceller. In our experiments, this system significantly reduced the errors in automatic speech recognition compared with the conventional combination of echo cancellation and spectral subtraction.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1093/ietfec/e88-a.7.1732/_p

Copy

@ARTICLE{e88-a_7_1732,
author={Osamu ICHIKAWA, Masafumi NISHIMURA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Simultaneous Adaptation of Echo Cancellation and Spectral Subtraction for In-Car Speech Recognition},
year={2005},
volume={E88-A},
number={7},
pages={1732-1738},
abstract={Recently, automatic speech recognition in a car has practical uses for applications like car-navigation and hands-free telephone dialers. For noise robustness, the current successes are based on the assumption that there is only a stationary cruising noise. Therefore, the recognition rate is greatly reduced when there is music or news coming from a radio or a CD player in the car. Since reference signals are available from such in-vehicle units, there is great hope that echo cancellers can eliminate the echo component in the observed noisy signals. However, previous research reported that the performance of an echo canceller is degraded in very noisy conditions. This implies it is desirable to combine the processes of echo cancellation and noise reduction. In this paper, we propose a system that uses echo cancellation and spectral subtraction simultaneously. A stationary noise component for spectral subtraction is estimated through the adaptation of an echo canceller. In our experiments, this system significantly reduced the errors in automatic speech recognition compared with the conventional combination of echo cancellation and spectral subtraction.},
keywords={},
doi={10.1093/ietfec/e88-a.7.1732},
ISSN={},
month={July},}

Copy

TY - JOUR
TI - Simultaneous Adaptation of Echo Cancellation and Spectral Subtraction for In-Car Speech Recognition
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1732
EP - 1738
AU - Osamu ICHIKAWA
AU - Masafumi NISHIMURA
PY - 2005
DO - 10.1093/ietfec/e88-a.7.1732
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E88-A
IS - 7
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - July 2005
AB - Recently, automatic speech recognition in a car has practical uses for applications like car-navigation and hands-free telephone dialers. For noise robustness, the current successes are based on the assumption that there is only a stationary cruising noise. Therefore, the recognition rate is greatly reduced when there is music or news coming from a radio or a CD player in the car. Since reference signals are available from such in-vehicle units, there is great hope that echo cancellers can eliminate the echo component in the observed noisy signals. However, previous research reported that the performance of an echo canceller is degraded in very noisy conditions. This implies it is desirable to combine the processes of echo cancellation and noise reduction. In this paper, we propose a system that uses echo cancellation and spectral subtraction simultaneously. A stationary noise component for spectral subtraction is estimated through the adaptation of an echo canceller. In our experiments, this system significantly reduced the errors in automatic speech recognition compared with the conventional combination of echo cancellation and spectral subtraction.
ER -