Litcius/Paper detail

A Robust Deep Audio Splicing Detection Method via Singularity Detection Feature

Kanghao Zhang, Shan Liang, Shuai Nie, Shulin He, Jiahui Pan, Xueliang Zhang, Haoxin Ma, Jiangyan Yi

2022ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)11 citationsDOI

Abstract

There are many methods for detecting forged audio produced by conversion and synthesis. However, as a simpler method of forgery, splicing has not attracted widespread attention. Based on the characteristic that the tampering operation will cause singularities at high-frequency components, we propose a high-frequency singularity detection feature obtained by wavelet transform. The proposed feature can explicitly show the location of the tampering operation on the waveform. Moreover, the long short-term memory (LSTM) is introduced to the CNN-architecture LCNN to ensure that the sequence information can be fully learned. The proposed feature is sent to the improved RNN-architecture LCNN together with the widely used linear frequency cepstral coefficients (LFCC) to learn forgery characteristics where the LFCC is used as a supplement. Systematic evaluation and comparison show that the proposed method has greatly improved the accuracy and generalization.

Topics & Concepts

Computer scienceFeature (linguistics)SingularityPattern recognition (psychology)Artificial intelligenceMel-frequency cepstrumGeneralizationCepstrumFeature extractionWaveletSpeech recognitionMathematicsMathematical analysisPhilosophyLinguisticsDigital Media Forensic DetectionMusic and Audio ProcessingSpeech and Audio Processing
A Robust Deep Audio Splicing Detection Method via Singularity Detection Feature | Litcius