PROJECT TITLE :

Anti-Forensics of Environmental-Signature-Based Audio Splicing Detection and Its Countermeasure via Rich-Features Classification

ABSTRACT:

Varied ways for detecting audio splicing are proposed. Environmental-signature-primarily based strategies are considered to be the most effective forgery detection ways. The performance of existing audio forensic analysis ways is mostly measured in the absence of any anti-forensic attack. Effectiveness of these methods within the presence of anti-forensic attacks is thus unknown. In this paper, we tend to propose an efficient anti-forensic attack for environmental-signature-based mostly splicing detection method and countermeasures to detect the presence of the anti-forensic attack. For anti-forensic attack, dereverberation-based processing is proposed. Three dereverberation methods are thought-about to tamper with the acoustic surroundings signature. Experimental results indicate that the proposed dereverberation-based mostly anti-forensic attack considerably degrades the performance of the selected splicing detection method. The proposed countermeasures exploit artifacts introduced by the anti-forensic processing. To detect the presence of potential anti-forensic processing, a machine learning-primarily based framework is proposed. In specific, the proposed anti-forensic detection method uses a rich-feature model consisting of Fourier coefficients, spectral properties, high-order statistics of musical noise residuals, and modulation spectral coefficients to capture traces of dereverberation attacks. The performance of the proposed framework is evaluated on both synthetic information and real-world speech recordings. The experimental results show that the proposed made-feature model will detect the presence of anti-forensic processing with a mean accuracy of 95%.