Paper Title:
Combining Speech Enhancement and Cepstral Mean Normalization for LPC Cepstral Coefficients
  Abstract

A mismatch between the training and testing in noisy circumstance often causes a drastic decrease in the performance of speech recognition system. The robust feature coefficients might suppress this sensitivity of mismatch during the recognition stage. In this paper, we investigate the noise robustness of LPC Cepstral Coefficients (LPCC) by using speech enhancement with feature post-processing. At front-end, speech enhancement in the wavelet domain is used to remove noise components from noisy signals. This enhanced processing adopts the combination of discrete wavelet transform (DWT), wavelet packet decomposition (WPD), multi-thresholds processing etc to obtain the estimated speech. The feature post-processing employs cepstral mean normalization (CMN) to compensate the signal distortion and residual noise of enhanced signals in the cepstral domain. The performance of digit speech recognition systems is evaluated under noisy environments based on NOISEX-92 database. The experimental results show that the presented method exhibits performance improvements in the adverse noise environment compared with the previous features.

  Info
Periodical
Key Engineering Materials (Volumes 474-476)
Edited by
Garry Zhu
Pages
349-354
DOI
10.4028/www.scientific.net/KEM.474-476.349
Citation
J. Yang, "Combining Speech Enhancement and Cepstral Mean Normalization for LPC Cepstral Coefficients", Key Engineering Materials, Vols. 474-476, pp. 349-354, 2011
Online since
April 2011
Authors
Export
Price
$32.00
Share

In order to see related information, you need to Login.

In order to see related information, you need to Login.

Authors: Wen Han Ma, Xiao Yi Hu, Li Yang Bai
Abstract:This paper describes the characteristics of the impulse noise in the shallow water acoustic channel, and its influence on the bit error rate...
1537
Authors: Zi Ran Liu, Tao He, Yuan Yuan He, Yu Xi Yu
Measure Control Technologies and Intelligent Systems
Abstract:Because of discontinuity at threshold, hard threshold de-noising leads to the additional oscillation at threshold when reconstructing the...
1201
Authors: Li Da Liao, Qing Hua He, Zhong Lin Hu
Chapter 3: Vibration Control and Condition Monitoring
Abstract:In order to identify noise sources of an excavator in non-library environment, a complex-valued algorithm in frequency domain was applied....
723
Authors: Cai Yun Wu
Chapter 1: Mechatronics
Abstract:This paper studies the speech enhancement technology. A Lyapunov function of the tracking error systems is defined, and the adaptive filter...
1670
Authors: Xin Xu
Chapter 11: Image Processing Technology
Abstract:One method is proposed to remove the random noise and low-frequency coherent noise in the images of the optic 4f system, which is based on...
2469