Calibrated Audio Steganalysis

TL;DR

This paper introduces a new calibrated feature set for audio steganalysis based on reembedding, significantly improving detection accuracy, especially at low embedding rates, by leveraging a model designed to deviate from human auditory perception.

Contribution

The paper proposes a novel set of calibrated features for audio steganalysis using reembedding, tailored to exploit deviations from human auditory perception, achieving higher accuracy than previous methods.

Findings

Achieves 99.3% accuracy at low embedding rates

Outperforms previous R-MFCC based methods by 9.5%

Effective in detecting hidden messages in audio signals

Abstract

Calibration is a common practice in image steganalysis for extracting prominent features. Based on the idea of reembedding, a new set of calibrated features for audio steganalysis applications are proposed. These features are extracted from a model that has maximum deviation from human auditory system and had been specifically designed for audio steganalysis. Ability of the proposed system is tested extensively. Simulations demonstrate that the proposed method can accurately detect the presence of hidden messages even in very low embedding rates. Proposed method achieves an accuracy of 99.3% (StegHide@0.76% BPB) which is 9.5% higher than the previous R-MFCC based steganalysis method.