Study on Characteristic Parameters of Speech Anti-Deliberate Imitation System

Abstract:

Deliberate imitation which is the reproduction of another speakers voice and speech behavior can pose a threat to the security of the voice authentication system. Therefore effective characteristic parameters are the key to the anti-deliberate imitation. The study chose speech database of anti-deliberate imitation and investigated some common feature parameters separating capacity and descriptive power against voice deliberate imitation. The study compared the ranking of subjective evaluation and feature parameters Euclidean distance of imitators. The comparison results indicate that Mel frequency cepstrum coefficient (MFCC) combined with its differential cepstrum parameters (WMFCC) have the best performance for anti-deliberate imitation. And the results were validated by the experiment based on VQ speaker verification system.

• Periodical: Applied Mechanics and Materials (Volumes 475-476)
• Pages: 388-393
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.475-476.388
• Online since: December 2013
• Authors: Ping Zhou, Jing Jing Ke*, Xin Xing Jing, Zhao Guo Cui
• Keywords: Characteristic Parameters, Deliberate Imitation, Euclidean Distance, Speaker Verification, Subjective Evaluation

References

[1] Xiuping Zhao, The development of biometric identification technology. Forensic Science and Technology, 2011(06): 44 -48. (In Chinese).

[2] W. Majewski, and P. Staroniewicz, Acoustical parameters of target voice and their imitators. Speech and Language Technology, 2008, 11: 17 -23.

[3] Shuzha Wang and Tianshuang. Qiu, Reviews on speaker recognition research. Audio Engineering, 2007, 31: 51 -55. (In Chinese).

[4] Lihua Mo, Xiaojin Zhou and Xiaojun Zhang, Study on pathological voice recognition based on LPCC and MFCC parameters. Communication Technology, 2012, 45: 87 -89. (In Chinese).

[5] Hong Liu and Liqun Liu, Study on improved mel frequency cepstrum coefficient for speaker recognition. Computer Engineering anf Applications, 2008, 48: 155 -157. (In Chinese).

DOI: 10.3724/sp.j.1087.2012.02542

[6] Ping Sun, Mingliang Gu and Biao Zhang, Dynamic time-frequency cepstrum of voice signals. Journal of Xuzhou Normal University (Natural Science Edition), 2011, 29: 58 -61. (In Chinese).

[7] Yan He and Fengqin Yu, A dialect recognition method based on static and dynamic features. Computer Engineering anf Applications, 2012, 48: 105 -108. (In Chinese).

[8] W. Majewski. Mel Frequency Cepstral Coefficients (MFCC) of original speakers and their imitators. Archives of Acoustics, 2006, 31: 445 -449.

[9] Weiwei Zhang, Study on speaker recognition technology. Yanshan University, Qinhuangdao, China. 2010. (In Chinese).

[10] Chaohui Wu, Speaker Recognition Models and Methods. Tsinghua University Publisher, Beijing, China. 2008. (In Chinese).

[11] Ziming Lin, Study of speaker recognition system based on GMM and VQ. South China University of Technology, Guangzhou, China. 2012. (In Chinese).