Unweighted fusion in microphone forensics using a decision tree and linear logistic regression models

Source microphone identification from speech recordings based on a Gaussian mixture model

“…It was reported in [11] that, using linear regression models, a 93.5% accurate classification rate was obtained. In another study [12], using a decision tree and linear logistic regression models, a 100% classification rate was obtained for 4 and 7 microphones. The results in [13] were obtained using MFCCs and linear scale cepstral coefficients with the SVM classifier.…”

“…They tested 7 different microphones and achieved a 93.5% correct classification rate with linear regression models. In [12], it was pointed out that fusion operations, such as the match level, rank level, and decision level, could be implemented for reliable microphone classification. Using the same microphones that were used in [10,11], 100% accuracy was reached via the method of rank level fusion.…”

“…Therefore, the main aim of the present study is to find out which of the several feature extraction techniques are best suited to source microphone identification systems. Although MFCCs have recently been exploited successfully in microphone forensics [10][11][12][13], our focus is on linear prediction cepstrum coefficients (LPCCs) and perceptuallybased linear predictive coefficients (PLPCs), thanks to their widespread use and easy modeling. Additionally, we aim to find out the important parameters that affect the performance of microphone recognition.…”

Acoustic environment identification using unsupervised learning

“…2. Statistical pattern recognition based techniques [9][10][11][12][13][14][15][16][17] have been proposed for recording location and device identification. However, these methods are limited by their low accuracy and inability to uniquely map an audio recording to the source.…”

“…Motivation behind considering acoustic artifacts for AEI and audio forensic applications is that existing audio forensic analysis methods, e.g., ENF-based methods [5][6][7][8] and recording device identification based methods [11][12][13][14] cannot withstand lossy compress attack, e.g., MP3 compression. In our recent work [18,32], we have shown that acoustic reverberations can survive lossy compression attack, which is one of the motivations behind considering acoustic artifacts in an audio recording for AEI and digital audio forensic applications.…”