Speaker-independent source cell-phone identification for re-compressed and noisy audio recordings

With the rapid increase in user-generated multimedia content, extensive outreach over social media, and their potential in critical applications such as law enforcement, sourcey identification from re-compressed and noisy multimedia are of great importance. This paper proposes a system for speaker-independent cell-phone identification from recorded audio. This system is capable of dealing with test audio with different speech content and a different speaker compared to the training audio. Each recorded audio has the device fingerprint implicitly embedded in it, which encourages us to design a CNN-based system for learning the device-specific signatures directly from the magnitude of discrete Fourier transform of the audio. This paper also addresses the scenario where the recorded audio is re-compressed due to efficient storage and network transmission requirements, which is a common phenomenon in this age of social media. The scenario of the cell-phone classification from the audio recordings in the presence of additive white Gaussian noise is addressed as well. We show that our proposed system performs as well as the state-of-art systems for the speaker-dependent case with clean audio recordings and exhibits much higher robustness in the speaker-independent case with clean, re-compressed, and noisy audio recordings.

[1]  Rangding Wang,et al.  Source Cell-Phone Identification in the Presence of Additive Noise from CQT Domain , 2018, Inf..

[2]  Tomi Kinnunen,et al.  Source cell-phone recognition from recorded speech using non-speech segments , 2014, Digit. Signal Process..

[3]  Junfeng Wu,et al.  Source cell phone verification from speech recordings using sparse representation , 2017, Digit. Signal Process..

[4]  Ömer Eskidere,et al.  Source microphone identification from speech recordings based on a Gaussian mixture model , 2014 .

[5]  Muhammad Khurram Khan,et al.  Digital multimedia audio forensics: past, present and future , 2017, Multimedia Tools and Applications.

[6]  Min Wu,et al.  Information Forensics: An Overview of the First Decade , 2013, IEEE Access.

[7]  Ömer Eskidere,et al.  Identifying acquisition devices from recorded speech signals using wavelet-based features , 2016 .

[8]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[9]  Yuhan Zhang,et al.  Mobile Phone Clustering From Speech Recordings Using Deep Representation and Spectral Clustering , 2018, IEEE Transactions on Information Forensics and Security.

[10]  Bin Li,et al.  Detection of Double Compressed AMR Audio Using Stacked Autoencoder , 2017, IEEE Transactions on Information Forensics and Security.

[11]  Gianmarco Baldini,et al.  Smartphones Identification Through the Built-In Microphones With Convolutional Neural Network , 2019, IEEE Access.

[12]  Cemal Hanilçi,et al.  Recognition of Brand and Models of Cell-Phones From Recorded Speech Signals , 2012, IEEE Transactions on Information Forensics and Security.

[13]  F.H.F. Leung,et al.  Source Microphone Recognition Aided by a Kernel-Based Projection Method , 2019, IEEE Transactions on Information Forensics and Security.

[14]  Jiwu Huang,et al.  Band Energy Difference for Source Attribution in Audio Forensics , 2018, IEEE Transactions on Information Forensics and Security.