Speaker recognition method based on deep residual network and improved power normalized Cepstral coefficients features

Runhua He; Pan Li; Xuemei Li; Shuhang Chen

doi:10.1117/12.2626663

16 December 2021 Speaker recognition method based on deep residual network and improved power normalized Cepstral coefficients features

Runhua He, Pan Li, Xuemei Li, Shuhang Chen

Author Affiliations +

Proceedings Volume 12153, International Conference on Artificial Intelligence, Virtual Reality, and Visualization (AIVRV 2021); 121530B (2021) https://doi.org/10.1117/12.2626663
Event: International Conference on Artificial Intelligence, Virtual Reality, and Visualization (AIVRV 2021), 2021, Sanya, China

Abstract

In recent years, speaker recognition technology can achieve more than 90% accuracy in quiet environment, but it is not accurate in noise environment. In order to solve the problem of speaker recognition in noisy environment, a speaker recognition method based on deep residual network and improved Power Normalized Cepstral Coefficients (PNCC) features is proposed in the study. The PNCC features extraction algorithm is improved to be suitable for PNCC features parameter extraction in noisy environment. In order to further improve the training effect of the model, the deep residual network is adopted in this study to train the model, which effectively improves the model accuracy. In the recognition stage, in order to effectively suppress the influence of noise, this study proposes a VMD algorithm based on wavelet threshold denoising to improve the accuracy of speaker recognition. The experimental results show that compared with MFCC and PNCC features parameter extraction methods, the improved PNCC features extraction method is more conducive to feature extraction of speech signal under noise environment. Compared with other speaker recognition methods, the proposed speaker recognition method based on deep residual network and improved PNCC features has higher recognition accuracy.

Citation Download Citation

Runhua He, Pan Li, Xuemei Li, and Shuhang Chen "Speaker recognition method based on deep residual network and improved power normalized Cepstral coefficients features", Proc. SPIE 12153, International Conference on Artificial Intelligence, Virtual Reality, and Visualization (AIVRV 2021), 121530B (16 December 2021); https://doi.org/10.1117/12.2626663

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
8 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Speaker recognition

Feature extraction

Detection and tracking algorithms

Signal processing

Wavelets

Denoising

Interference (communication)

Show All Keywords

Keywords/Phrases

Search In:

Publication Years