Detecting Steganography of Adaptive Multirate Speech with Unknown Embedding Rate

Abstract

Steganalysis of adaptive multirate (AMR) speech is a significant research topic for preventing cybercrimes based on steganography in mobile speech services. Differing from the state-of-the-art works, this paper focuses on steganalysis of AMR speech with unknown embedding rate, where we present three schemes based on support-vector-machine to address the concern. The first two schemes evolve from the existing image steganalysis schemes, which adopt different global classifiers. One is trained on a comprehensive speech sample set including original samples and steganographic samples with various embedding rates, while the other is trained on a particular speech sample set containing original samples and steganographic samples with uniform distributions of embedded information. Further, we present a hybrid steganalysis scheme, which employs Dempster-Shafer theory (DST) to fuse all the evidence from multiple specific classifiers and provide a synthesized detection result. All the steganalysis schemes are evaluated using the well-selected feature set based on statistical characteristics of pulse pairs and compared with the optimal steganalysis that adopts specialized classifiers for corresponding embedding rates. The experimental results demonstrate that all the three steganalysis schemes are feasible and effective for detecting the existing steganographic methods with unknown embedding rates in AMR speech streams, while the DST-based scheme outperforms the others overall.

Steganography and Steganalysis in Voice over IP: A Review

“…Huang et al [75], in 2017, proposed a hybrid steganalysis scheme. The pulse pair features are extracted after grouping and processing the training samples, and they are trained separately by specific classifiers.…”

A Common Method for Detecting Multiple Steganographies in Low-Bit-Rate Compressed Speech Based on Bayesian Inference

“…Since PSPP features describe only the distributions of pulses being in the same position in two tracks, Tian et al [38] employ probability distributions, Markov transition probabilities and joint probabilities to characterize pulse pairs and use the adaptive boosting technique to reduce the feature dimensionality. Based on the above features, they propose three classification schemes to achieve steganalysis for unknown embedding rates in [39]. Li et al [40] propose a steganalysis method for pitch period modification steganography based on a codebook correlation network model, in which the conditional probabilities of strongly correlated nodes are used as features and principal component analysis (PCA) is applied for dimensionality reduction before training the SVM classifier.…”

A new approach for enhancing LSB steganography using bidirectional coding scheme

“…The term steganalysis refers to techniques used to reveal the hidden information embedded in the cover file [15,16]. Many algorithms were introduced for this purpose.…”