Abstract
The problem of the existing channel impulse response based audio splicing tampering passive forensics methods is the low detection rate. Therefore, in this paper an audio splicing detection and localization method based on channel response multi-feature is proposed. Firstly, we use Gaussian Mixture Model (GMM) on RASTA-MFCC coefficients to blind channel estimation. Then use channel response of the audio signal and dynamic and static information of the logarithmic spectral characteristics for the query signal to construct the feature to identify the tampering and locate. We use sequence forward selection (SFS) algorithm for filtering out the features. At the time of location detection, we first determine whether the audio is tampered and then start with a number of audio frame features average value that serve as a reference feature. The correlation coefficients of reference features and each of other audio frame features are calculated separately. On comparison with the existing algorithm, the results show the algorithm improves the detection rate of forgery and its localization.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
