التفاصيل البيبلوغرافية
العنوان: |
Noise-robust feature based on sparse representation for speaker recognition. |
المؤلفون: |
Hongzhuo Qi |
المصدر: |
Metallurgical & Mining Industry; 2015, Issue 4, p64-69, 6p |
مصطلحات موضوعية: |
VOICEPRINTS, NOISE (Work environment), ROBUST control |
مستخلص: |
The performance of speaker recognition suffers substantial degradation in noisy environments. To solve this problem, we propose a new feature extraction method based on sparse coding to improve the noise robustness of the speaker recognition. In this method, an over-complete dictionary, which is powerful in identifying transient underlying structures and harmonic periodicities, is trained with amounts of unlabeled data. Next, speech signals are sparsely represented by atoms of the dictionary. After that, the proposed feature is extracted from the sparse representation by a tuning function which simulates the hearing mechanism. Experiments show that the proposed method outperforms the mel-frequency cepstral coefficients (MFCC) and the perceptual linear predictive (PLP) feature in various noise conditions. [ABSTRACT FROM AUTHOR] |
|
Copyright of Metallurgical & Mining Industry is the property of Ukrmetallurginform STA Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) |
قاعدة البيانات: |
Complementary Index |