Abstract:The cover source mismatch is a common problem in steganalysis, which may result in the degradation of detection accuracy. In this paper, we present a novel method to mitigate the problem of JPEG quantization table mismatch, named as Robust Discriminative Feature Transformation (RDFT). RDFT transforms original features to new feature representations based on a non-linear transformation matrix. It can improve the statistical consistency of the training samples and testing samples and learn new matched feature representations from original features by minimizing feature distribution difference while preserving the classification ability of training data. The comparison to prior arts reveals that the detection accuracy of the proposed RDFT algorithm can significantly outperform traditional steganalyzers under mismatched conditions and it is close to that of matched scenario. RDFT has several appealing advantages: 1) it can improve the statistical consistency of the training and testing data; 2) it can reduce the distribution difference between the training features and testing features; 3) it can preserve the classification ability of the training data; 4) it is robust to parameters and can achieve a good performance under a wide range of parameter values.

A JPEG Steganalysis Algorithm

Jpeg Quantization-Distribution Steganalytic Method Attacking Jsteg

Feature-Based Steganalysis for JPEG Images

Steganalysis of model based steganography and steghide in grayscale JPEG images

JPEG Steganalysis Based on Co-occurrence Features and Ensemble Multiple Hyperspheres OC-SVM

Steganalytic Scheme for JPEG Images Based on Multivariate Regression

A New Steganographic Method Resisting Statistical Attack

Multi-class Blind Steganalysis for JPEG Images Based on Micro-templates

Double-Compression Detection in JPEG Steganalysis

Steganalysis based on differential statistics

A JPEG Steganalysis Algorithm with Embedded Rate Estimation Based on Multi-class SVM

Novel Blind Steganalysis For Jpeg Images

Double-compressed JPEG images steganalysis with transferring feature

A Blind Steganalytic Scheme Based on DCT and Spatial Domain for JPEG Images.

A co-occurrence matrix based approach to detect jpeg steganography

Blind Jpeg Steganalysis Using Features Derived from Multi-Domain

A New Steganalysis Approach Based on Both Complexity Estimate and Statistical Filter

JPEG Quantization Table Mismatched Steganalysis Via Robust Discriminative Feature Transformation.

Blind detection for JPEG steganography

Variable Characteristics Based Blind Detection of Hidden Information

Blind JPEG steganalysis based on multi-domain features