The State of Art in Digital Audio Forensic Authentication
Zeng Jin-hu · Zhongguo sifa jianding · 2014
With the popularity of digital recorders, and the popular trend of audio tampering methods, the traditional techniques have met great challenges in the current practice of digital audio forensic authentication. The latest progress in the field of the pattern recognition and artificial intelligence offers a feasible way for digital audio forensic authentication. In this paper,the latest achievements of digital audio forensic authentication in the fields of the machine learning and pattern recognition were reviewed and the current main methods in the practice of digital audio forensic authentication were analyzed. Then, problems, challenges, and future trends of digital audio forensic authentication were discussed. It was concluded that the collaborative coexistence of traditional methods based on expert experience and statistical quantitative computation techniques would be an inevitable trend and an efficient solution for digital audio forensic authentication.