In a real-world cloud server, a speech signal is prone to suffer various attacks, such as malicious muting and tampering. In such a context, the privacy security of the speech owner will not be guaranteed. In order to achieve content authentication of encrypted speech in the cloud server, an efficient encrypted speech authentication method based on uniform subband spectrum variance and perceptual hashing is proposed. Firstly, the original speech is scrambled by Henon mapping to construct an encrypted speech library in the cloud, through extracting uniform subband spectrum variance of the encrypted speech and constructing a hashing sequence to generate a hashing template of the cloud. In this way, a one-to-one correspondence between the encrypted speech and the hashing sequence is built. Secondly, the authentication digest of encrypted speech is extracted according to the inquiry result. Finally, the authentication digest and the hashing sequence in the cloud are matched by the Hamming distance algorithm. The experimental results demonstrate that the proposed method has great security and efficiency, and it can directly extract the authentication digest from encrypted speech. The authentication digest not only has good discrimination and robustness, but it accurately locates the tampered area for malicious substitution and mute attacks.
Encrypted speech authentication, perceptual hashing, uniform subband spectrum variance, feature extraction of encrypted speech, tamper location
ZHANG, QIUYU; ZHANG, DENGHAI; and ZHOU, LIANG
"An encrypted speech authentication method based on uniform subband spectrumvariance and perceptual hashing,"
Turkish Journal of Electrical Engineering and Computer Sciences: Vol. 28:
5, Article 7.
Available at: https://journals.tubitak.gov.tr/elektrik/vol28/iss5/7