Feature-Based Steganalysis for JPEG Images and Its Implications for Future Design of Steganographic Schemes

By Jessica J. Fridrich. Information Hiding, 2004

584 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/dd54fd11b782d5a4bd68cb22be7338dfa2315c00

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Jessica J. Fridrich
2. Vojtech Holub
3. Bin Li
4. Jan Kodovský
5. Tomás Pevný
6. Y. Shi
7. Jiwu Huang
8. Junhui He

In Computer Vision and Pattern Recognition

1. Jessica J. Fridrich
2. Vojtech Holub
3. Jan Kodovský
4. Tomás Pevný
5. Bin Li
6. Jiwu Huang
7. Luisa Verdoliva
8. Y. Shi

Citations in the corpus

• Steganalysis of LSB matching using differences between nonadjacent pixels (2016, methodology): “In this paper, the detection performance of steganalysis method is evaluated by the minimized classification error defined as [8]:”
• Benchmarking for Steganography (2008, background): “), and a functional assigning a single numerical value to the ROC curve (total minimal decision error [16, 38], probability of detection for fixed false alarm rate [26], or false alarm for probability of detection 50% [20], accuracy [13], etc.”
• Ensemble Classifiers for Steganalysis of Digital Media (2012, methodology): “…used only a few dozen features, e.g., 72 higher order moments of coefficients obtained by transforming an image using quadratic mirror filters [14], 18 binary similarity metrics [3], 23 discrete cosine transform (DCT) features [17], and 27 higher order moments of wavelet coefficients [21].”
• Depth-Wise Separable Convolutions and Multi-Level Pooling for an Efficient Spatial CNN-Based Steganalysis (2020, methodology): “The commonly used statistical features include the binary similarity measure feature [7], DCT [8], [9] and wavelet coefficient feature [10], co-occurrence matrix feature [11] and so on.”
• Statistically undetectable jpeg steganography: dead ends challenges, and opportunities (2007, methodology): “Based on the comparison in [29], blind steganalysis constructed using this feature vector achieves performance that is one of the best compared to currently available blind steganalysis classifiers capable of detecting stego content in JPEG images [3, 2, 33, 1, 25, 10, 39, 9, 29, 12].”
• Steganalysis of LSB matching in grayscale images (2005, background): “Finally, we outline some further directions for research.”
• Detection of Double-Compression in JPEG Images for Applications in Steganography (2008, methodology): “estimates obtained by calibration [4], [5] followed by simulated double-compression.”
• An improved approach to steganalysis of JPEG images (2010, methodology): “[23] T. Pevny, J. Fridrich, Merging Markov and DCT features for multi-class JPEG steganalysis, in: Proceedings of SPIE Electronic Imaging, Electronic Imaging, Security, Steganography, and Watermarking of Multimedia Contents IX, vol. 6505, 2007.”
• Steganalysis of least significant bit matching using multi-order differences (2014, background): “Detection performance is evaluated by “detection reliability” ρ [22] defined as”
• A Markov Process Based Approach to Effective Attacking JPEG Steganography (2006, background): “It appears contradicting to what reported in [7, 16].”
• Steganalysis using higher-order image statistics (2006, background): “Current steganalysis methods fall broadly into one of two categories: embedding specific (e.g., [7]–[12]) or universal (e.g., [13]–[17]).”
• Steganalysis of JPEG images using rich models (2012, background): “Current trends in steganalysis Modern steganography requires more complex feature spaces • 18 BSM [Avcıbaş,2002], 72 higher–order moments [Farid,2002] • 23 DCT [Fridrich,2004]→ 274 PEV [Pevný,2007]→ 548 CC-PEV • 324 [Shi,2006]→ 486 [Chen,2008] – Markov–based features • 686 SPAM [Pevný,2010]→ 1234 CD”
• A Survey on Image Steganography and Steganalysis (2011, methodology): “As discussed in Section 4.2.2, Fridrich [34] designed a powerful calibration method for JPEG images through cropping and re-compressing.”
• Merging Markov and DCT features for multi-class JPEG steganalysis (2007, methodology): “4,5,9,10 Such steganalytic methods are called blind.”
• Rich Models for Steganalysis of Digital Images (2012, background): “The model is usually determined not only by the characteristics of the cover source but also by the effects of embedding [4], [6], [14], [18], [21], [33], [35], [38].”