Detection of audio copy-move-forgery with novel feature matching on Mel spectrogram

作者：

Highlights：

• Audio copy-move forgery detection and localization.

• Detect forgeries even under post-processing operations applied to forged speech to hide traces of forgery.

• Using a keypoint-based approach on the Mel spectrogram representation of audio.

• Performance against common post-processing operations such as noise addition, filtering operation, and especially compression operation.

摘要

•Audio copy-move forgery detection and localization.•Detect forgeries even under post-processing operations applied to forged speech to hide traces of forgery.•Using a keypoint-based approach on the Mel spectrogram representation of audio.•Performance against common post-processing operations such as noise addition, filtering operation, and especially compression operation.

论文关键词：Audio copy-move-forgery detection,Audio forgery,Audio forensic,Mel spectrogram,SIFT keypoints,Forgery localization

论文评审过程：Received 12 June 2022, Revised 26 September 2022, Accepted 1 October 2022, Available online 12 October 2022, Version of Record 17 October 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118963