Digital Audio Tampering Detection Based on ENF Consistency

This paper addresses a method of automatic detection of digital audio signal tampering based on feature fusion. Aiming at the insertion and deletion operations in the digital audio signal tamper chain. In this paper, the Electric Network Frequency (ENF) component of the digital audio signal is extracted and the consistency of the ENF component is analyzed to determine whether the audio signal is tampered with. In this paper, a general framework for passive tamper detection of audio signal based on ENF component consistency and a general framework for ENFC feature extraction are proposed. The feature set is used to quantify the amplitude of the phase and instantaneous frequency variations of the ENF component and to serve as an indicator of the consistency of the ENF component. SVM classifier is used to classify the extracted feature sets. The experimental results show that this method can classify the original signal and the edit signal which is inserted and deleted.

[1]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[2]  M. D. Catherine.  High precision Fourier analysis of sounds using signal derivatives , 2000 .

[3]  José Antonio Apolinário,et al.  Edit Detection in Speech Recordings via Instantaneous Electric Network Frequency Variations , 2014, IEEE Transactions on Information Forensics and Security.

[4]  Leon Cohen,et al.  Time Frequency Analysis: Theory and Applications , 1994 .

[5]  Yilu Liu,et al.  Application of Power System Frequency for Digital Audio Authentication , 2012, IEEE Transactions on Power Delivery.

[6]  Catalin Grigoras.  Applications of ENF criterion in forensic audio, video, computer and telecommunication analysis. , 2007, Forensic science international.

[7]  Daniel Patricio Nicolalde Rodríguez,et al.  Evaluating digital audio authenticity with spectral distances and ENF phase change , 2009, 2009 IEEE International Conference on Acoustics, Speech and Signal Processing.

[8]  Daniel Patricio Nicolalde Rodríguez,et al.  Audio Authenticity: Detecting ENF Discontinuity With High Precision Phase Analysis , 2010, IEEE Transactions on Information Forensics and Security.

[9]  Catalin Grigoras.  Digital audio recording analysis: the Electric Network Frequency (ENF) Criterion , 2005 .

[10]  Sylvain Marchand,et al.  High-Precision Fourier Analysis of Sounds Using Signal Derivatives , 2000 .

[11]  Eftim Zdravevski,et al.  SVM Parameter Tuning with Grid Search and Its Impact on Reduction of Model Over-fitting , 2015, RSFDGrC.

[12]  Bernhard Schölkopf,et al.  Comparing support vector machines with Gaussian kernels to radial basis function classifiers , 1997, IEEE Trans. Signal Process..