An Intriguing Struggle of CNNs in JPEG Steganalysis and the OneHot Solution

Yousfi, Yassine; Fridrich, Jessica

doi:10.1109/lsp.2020.2993959

Cited by 43 publications

(17 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We only show the cases when 93 ≤ Q 1 ≤ Q 2 and also when Q 1 , Q 2 ∈ {99, 100}, since these cases satisfy condition [C1]. Three detectors are tested: SRNet [1] trained on the rounding errors after decompressing the JPEG image (e-SRNet), JRM with the ensemble classifier [5], and OneHot network [9] combined with e-SRNet (eOH-SRNet), which is implemented as OneHot-SRNet in the original paper with clipping threshold T = 5. The SRNet, however, takes the rounding errors on the input instead of the spatial representation of the image.…”

Section: Resultsmentioning

confidence: 99%

“…We want to point out that both network based detectors converge to their optimum extremely quickly, within 20k iterations. Even though e-SRNet fails for some combinations of the compression qualities, such as (96, 97), double compression with such combinations of quality factors leads to peaks and valleys in cover DCT histograms, which allows very accurate detection with JRM and other prior art [6,8,10,9]. Note that these detectors perform rather poorly whenever Q 1 = Q 2 .…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Extending the Reverse JPEG Compatibility Attack to Double Compressed Images

Butora

Fridrich

2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

The reverse JPEG compatibility attack has recently been introduced as a very accurate and universal steganalysis algorithm for JPEG images with quality 99 or 100. The limitation to these two largest qualities appears fundamental as the prior work on this topic suggests. In this paper, we provide mathematical analysis and demonstrate experimentally that this attack can be extended to double compressed images when the first compression quality is 93 or larger and the second quality equal or larger than the first quality. Comparisons with stateof-the-art deep convolutional neural networks as well as detectors built in the JPEG domain show the merit of this work.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Extending the Reverse JPEG Compatibility Attack to Double Compressed Images

Butora

Fridrich

2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

show abstract

“…For this reason it has been quite unused. However, among those who did use DCT coefficients, undoubtedly the most successful approach has been to use the one hot encoding approach recently proposed in [21].…”

Section: Discussionmentioning

confidence: 99%

ALASKA#2: Challenging Academic Research on Steganalysis with Realistic Images

Cogranne

Giboulot

Bas

2020

2020 IEEE International Workshop on Information Forensics and Security (WIFS)

View full text Add to dashboard Cite

This paper briefly summarizes the ALASKA#2 steganalysis challenge which has been organized on the Kaggle machine learning competition platform. We especially focus on the context, the organization (rules, timeline, evaluation and material) as well as on the outcome (number of competitors, submission, findings, and final results). While both steganography and steganalysis were new to most of the competitors, they were able to leverage their skills in Deep Learning in order to design detection methods that perform significantly better than current art in steganalysis. Despite the fact that these solutions come at an important computational cost, they clearly indicate new directions to explore in steganalysis research.

show abstract

“…The model is tested against the SRnet, the proposed model with average pooling instead of Global covariance pooling. The authors in [137] presented OneHot CNN architecture to effectively detect the stego images in JPEG domain. The OneHot encodes the DCT coefficients of the images into binary volumetric representation of the DCT plane.…”

Section: Deep Learning Models For Image Steganalysismentioning

confidence: 99%

Stegomalware: A Systematic Survey of MalwareHiding and Detection in Images, Machine LearningModels and Research Challenges

Chaganti¹,

Sowmya²,

Alazab³

et al. 2021

Preprint

View full text Add to dashboard Cite

Malware distribution to the victim network is commonly performed through file attachments in phishing email or downloading illegitimate files from the internet, when the victim interacts with the source of infection. To detect and prevent the malware distribution in the victim machine, the existing end device security applications may leverage sophisticated techniques such as signature-based or anomaly-based, machine learning techniques. The well-known file formats Portable Executable (PE) for Windows and Executable and Linkable Format (ELF) for Linux based operating system are used for malware analysis and the malware detection capabilities of these files has been well advanced for real time detection. But the malware payload hiding in multimedia like cover images using steganography detection has been a challenge for enterprises, as these are rarely seen and usually act as a stager in sophisticated attacks. In this article, to our knowledge, we are the first to try to address the knowledge gap between the current progress in image steganography and steganalysis academic research focusing on data hiding and the review of the stegomalware (malware payload hiding in images) targeting enterprises with cyberattacks current status. We present the stegomalware history, generation tools, file format specification description. Based on our findings, we perform the detail review of the image steganography techniques including the recent Generative Adversarial Networks (GAN) based models and the image steganalysis methods including the Deep Learning (DL) models for hiding data detection. Additionally, the stegomalware detection framework for enterprise is proposed for anomaly based stegomalware detection emphasizing the architecture details for different network environments. Finally, the research opportunities and challenges in stegomalware generation and detection are presented based on our findings.

show abstract

An Intriguing Struggle of CNNs in JPEG Steganalysis and the OneHot Solution

Cited by 43 publications

References 21 publications

Extending the Reverse JPEG Compatibility Attack to Double Compressed Images

Extending the Reverse JPEG Compatibility Attack to Double Compressed Images

ALASKA#2: Challenging Academic Research on Steganalysis with Realistic Images

Stegomalware: A Systematic Survey of MalwareHiding and Detection in Images, Machine LearningModels and Research Challenges

Contact Info

Product

Resources

About