Steganalysis Over Large-Scale Social Networks With High-Order Joint Features and Clustering Ensembles

Li, Fengyong; Wu, Kui; Lei, Jingsheng; Wen, Mei; Bi, Zhongqin; Gu, Chunhua

doi:10.1109/tifs.2015.2496910

Cited by 49 publications

(21 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result, the trained model may not match well the features of test images, leading to poor detection results. To avoid the problem, steganalysts try to use cluster analysis [14,17] or anomaly detection [15,16] to solve this problem.…”

Section: Steganographer Detectionmentioning

confidence: 99%

“…On the basis of Ker's schemes, Li et al, presented a new solution [17] by using high-order joint features and ensemble scheme to further improved the performance of steganographer detection. High-order joint matrices of DCT coefficients of JPEG images are built to present new steganalysis feature set.…”

Section: Steganographer Detectionmentioning

confidence: 99%

“…As such, we need to detect which users are suspicious of using steganography to deliver hidden information. This new problem is termed as the steganographer detection problem and has been recently studied in [14][15][16][17].…”

Section: Introductionmentioning

confidence: 99%

“…Unsupervised learning is used to train a classifier which can separate "outliers" as steganographer from multiple innocent ones [14][15][16][17]. It has been shown that all the three schemes for the stego detection problem do not work well for steganographer detection [16], and only Ker's method [14][15][16] and Li's method [17] are applicable. Unfortunately, existing solutions to the steganographer detection problem can only identify a potential steganographer but cannot pin down to the steganographic images from the steganographer.…”

Section: Introductionmentioning

confidence: 99%

“…One natural question arises: could we adopt a two-step procedure to effectively identify the steganographic images when the source images are from many users? The two-step procedure first uses steganographer detection [14][15][16][17] to find potential steganographers and then applies existing stego detection methods to discover steganographic images.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Multi-Source Stego Detection with Low-Dimensional Textural Feature and Clustering Ensembles

Zhang

et al. 2018

Symmetry

Self Cite

View full text Add to dashboard Cite

This work tackles a recent challenge in digital image processing: how to identify the steganographic images from a steganographer, who is unknown among multiple innocent actors. The method does not need a large number of samples to train classification model, and thus it is significantly different from the traditional steganalysis. The proposed scheme consists of textural features and clustering ensembles. Local ternary patterns (LTP) are employed to design low-dimensional textural features which are considered to be more sensitive to steganographic changes in texture regions of image. Furthermore, we use the extracted low-dimensional textural features to train a number of hierarchical clustering results, which are integrated as an ensemble based on the majority voting strategy. Finally, the ensemble is used to make optimal decision for suspected image. Extensive experiments show that the proposed scheme is effective and efficient and outperforms the state-of-the-art steganalysis methods with an average gain from 4% to 6%.

show abstract