Audio tampering detection via microphone classification

Cuccovillo, Luca; Mann, Sebastian; Tagliasacchi, Marco; Aichroth, Patrick

doi:10.1109/mmsp.2013.6659284

Cited by 54 publications

(19 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The device information in the audio can be analyzed to determine whether the audio has been edited [7]. Cuccovillo et al [8] analyzes the microphones of recorded audio to detect the presence of multiple microphones in single audio for tampering detection. When recording audio, the surrounding environment mainly contains background noise, which can be used for audio tampering detection by analyzing background noise.…”

Section: Detection Methods Based On Shallow Featuresmentioning

confidence: 99%

Shallow and Deep Feature Fusion for Digital Audio Tampering Detection

Zeng¹,

Yang²,

Wang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Digital Audio tampering detection can be applied to verify the authenticity of digital audio. However, the current methods are mostly based on visual comparison analysis of the continuity of electronic network frequency (ENF) of digital audio with a standard ENF database. It is usually tricky to obtain the ENF database, and the feature expression of the visualization method is weak, which leads to low detection accuracy. In order to solve this problem, this paper proposed an audio tampering detection method based on the fusion of shallow and deep features. Firstly, the band-pass filtering process is performed on the audio signal to obtain the ENF components, and then the discrete Fourier transform and Hilbert transform are applied to obtain the phase and instantaneous frequency of the ENF components. Secondly, the shallow features are extracted by performing framing and fitting operations on the estimated phase and instantaneous frequency. Then, the designed convolutional neural network is used to obtain deep features, and the attention mechanism is applied to fuse shallow features and deep features. Finally, after dimensionality reduction through the fully connected layer, the Softmax layer is used for classification to detect the tampering audio. The method achieves 97.03% accuracy on three classic databases, which are Carioca 1, Carioca 2, and New Spanish. In addition, we have achieved an accuracy of 88.31% on the newly constructed database GAUDI-DI. Experimental results show that the proposed method is superior to the state-of-the-art method.

show abstract

Section: Detection Methods Based On Shallow Featuresmentioning

confidence: 99%

Shallow and Deep Feature Fusion for Digital Audio Tampering Detection

Zeng¹,

Yang²,

Wang³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…To prevent or detect tampering, digital signatures [4], information hiding [1], [2], [3], [5], [6], [7], [8], [9], [10], [11], [12], and other alternatives based on noise, device, and environment identification [13], [14], [15] iechizen@nii.ac.jp are widely used. In digital signature systems, a digital signature is appended to the header of the content and once it is removed, the content can no longer be verified.…”

Section: Introductionmentioning

confidence: 99%

“…In digital signature systems, a digital signature is appended to the header of the content and once it is removed, the content can no longer be verified. Technologies [13], [14], [15] can be used to detect whether material has been spliced with recordings from different acoustic environments or recording devices. In information hiding methods, the information for verification is embedded within the content itself.…”

Section: Introductionmentioning

confidence: 99%

Reversible Audio Information Hiding for Tampering Detection and Localization Using Sample Scanning Method

Huang

Nishimura

Echizen

2017

Journal of Information Processing

View full text Add to dashboard Cite

Reversible audio information hiding and sample-scanning methods are proposed for digital audio content to achieve detailed detection and localization of tampered positions in each frame. The method proposed in this study allows detecting multiple tampering and reusing reliable content as well as avoiding false detection which were impossible for other methods to simultaneously achieve. In the proposed method, the original signal is partitioned into fixed-length frames and then transformed into discrete cosine transform (DCT) coefficients by the integer modified DCT (intDCT). Expansion of the DCT coefficients is applied to embed a content-based hash as a payload. The integer DCT algorithm ensures the reversibility of the transform so that the original data and embedded payload can be perfectly restored to enable blind verification of the data integrity. The perceptual evaluation of speech quality (PESQ) with the listening quality objective mean opinion (MOSLQO), the segmental signal to noise ratio (segSNR), and subjective evaluation results show that the proposed algorithm provides good sound quality (MOSLQO and segSNR are respectively 4.41 and 23.31 dB on average for a capacity of 8,000 bps). Detection and localization are accurate in terms of correctly localizing tampered frames in case of insertion or deletion.

show abstract

“…Many people have done the identification of audio recording devices for numerous proposes in numerous conditions. Luca Cuccovillo et al [1] used microphone classification to perform audio tampering detection, and the underlying algorithm was based on blind channel estimation and applied to detect a specific type of tampering. Constantine Kotropoulos et al [2] performed research on mobile phone identification using recorded speech signals, and they used Mel frequency cepstral coefficients extracted from recorded speech signals to train a Gaussian Mixture Model with diagonal covariance matrices, thus providing templates for each device.…”

Section: Introductionmentioning

confidence: 99%

Audio recording device identification based on deep learning

Huang

et al. 2016

2016 IEEE International Conference on Signal and Image Processing (ICSIP)

View full text Add to dashboard Cite

In this paper we present a research on identification of audio recording devices from background noise, thus providing a method for forensics. The audio signal is the sum of speech signal and noise signal. Usually, people pay more attention to speech signal, because it carries the information to deliver. So a great amount of researches have been dedicated to getting higher Signal-Noise-Ratio (SNR). There are many speech enhancement algorithms to improve the quality of the speech, which can be seen as reducing the noise. However, noises can be regarded as the intrinsic fingerprint traces of an audio recording device. These digital traces can be characterized and identified by new machine learning techniques. Therefore, in our research, we use the noise as the intrinsic features. As for the identification, multiple classifiers of deep learning methods are used and compared. The identification result shows that the method of getting feature vector from the noise of each device and identifying them with deep learning techniques is viable, and well-preformed.

show abstract

Audio tampering detection via microphone classification

Cited by 54 publications

References 5 publications

Shallow and Deep Feature Fusion for Digital Audio Tampering Detection

Shallow and Deep Feature Fusion for Digital Audio Tampering Detection

Reversible Audio Information Hiding for Tampering Detection and Localization Using Sample Scanning Method

Audio recording device identification based on deep learning

Contact Info

Product

Resources

About