JPEG Steganalysis Detectors Scalable With Respect to Compression Quality

Yousfi, Yassine; Fridrich, Jessica

doi:10.2352/issn.2470-1173.2020.4.mwsf-075

Cited by 12 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both models were trained on all quality factors without the PC without any need to tweak the training schedule. While this is not surprising for B0 [29], it drastically simplifies the training of the SRNet across multiple quality factors [30]. As reported in [29], when training the SRNet from scratch, it had to be trained first on QF 75 with the PC to allow convergence, then refined without the PC, and finally trained via curriculum learning for the other two qualities.…”

Section: Jpeg Domainmentioning

confidence: 98%

How to Pretrain for Steganalysis

Butora

Yousfi

Fridrich

2021

Proceedings of the 2021 ACM Workshop on Information Hiding and Multimedia Security

View full text Add to dashboard Cite

In this paper, we investigate the effect of pretraining CNNs on Ima-geNet on their performance when refined for steganalysis of digital images. In many cases, it seems that just 'seeing' a large number of images helps with the convergence of the network during the refinement no matter what the pretraining task is. To achieve the best performance, the pretraining task should be related to steganalysis, even if it is done on a completely mismatched cover and stego datasets. Furthermore, the pretraining does not need to be carried out for very long and can be done with limited computational resources. An additional advantage of the pretraining is that it is done on color images and can later be applied for steganalysis of color and grayscale images while still having on-par or better performance than detectors trained specifically for a given source. The refining process is also much faster than training the network from scratch. The most surprising part of the paper is that networks pretrained on JPEG images are a good starting point for spatial domain steganalysis as well.

show abstract

Section: Jpeg Domainmentioning

confidence: 98%

How to Pretrain for Steganalysis

Butora

Yousfi

Fridrich

2021

Proceedings of the 2021 ACM Workshop on Information Hiding and Multimedia Security

View full text Add to dashboard Cite

show abstract

“…With this information, it is easy to identify whether the matrices are standard or non-standard. As recalled in the article of Yousfi and Fridrich [19], the standard formula for obtaining a quantization matrix whatever the channel, given the Quality Factor, Q, is:…”

Section: Reflection About Quantization Matrix Diversitymentioning

confidence: 99%

“…To do that, on can first estimate the non standard Quality Factor Q (ns) of the input JPEG image (trough iterative tests using the distance defined in Equation 8 of [19]), then compute the q (ns) (50) by multiplying the quantization matrices by the pre-computed "passage"matrices, from q(Q (ns) ) to q(50), and finally re-use equation 1 with substituting q(50) by q (ns) (50), this in order to obtain q (ns) (75).…”

Section: Reflection About Quantization Matrix Diversitymentioning

confidence: 99%

LSSD: a Controlled Large JPEG Image Database for Deep-Learning-based Steganalysis "into the Wild"

Ruiz¹,

Yedroudj²,

Chaumont³

et al. 2021

Preprint

View full text Add to dashboard Cite

For many years, the image databases used in steganalysis have been relatively small, i.e. about ten thousand images. This limits the diversity of images and thus prevents large-scale analysis of steganalysis algorithms. In this paper, we describe a large JPEG database composed of 2 million colour and grey-scale images. This database, named LSSD for Large Scale Steganalysis Database, was obtained thanks to the intensive use of "controlled" development procedures. LSSD has been made publicly available, and we aspire it could be used by the steganalysis community for large-scale experiments. We introduce the pipeline used for building various image database versions. We detail the general methodology that can be used to redevelop the entire database and increase even more the diversity. We also discuss computational cost and storage cost in order to develop images.

show abstract

“…Nevertheless, the conclusions obtained in the following could probably extend to a small interval around quality factor 75. Indeed, the authors of [28] show that applying a steganalysis with a JPEG images database made of multiple quality factors could be done without efficiency reduction, using a small set of dedicated networks, each targeting a small interval of quality factor. Finally, it is maybe possible to use a unique network when there are various quality factors, as it has been done by a majority of competitors during Alaska#2, thanks to the use of a pre-learned network on ImageNet.…”

Section: A Test Bench To Assess Scalability For Dl-based Steganalysismentioning

confidence: 99%

Analysis of the Scalability of a Deep-Learning Network for Steganography "Into the Wild"

Ruiz¹,

Chaumont²,

Yedroudj³

et al. 2020

Preprint

View full text Add to dashboard Cite

Since the emergence of deep learning and its adoption in steganalysis fields, most of the reference articles kept using small to medium size CNN, and learn them on relatively small databases. Therefore, benchmarks and comparisons between different deep learningbased steganalysis algorithms, more precisely CNNs, are thus made on small to medium databases. This is performed without knowing:1. if the ranking, with a criterion such as accuracy, is always the same when the database is larger, 2. if the efficiency of CNNs will collapse or not if the training database is a multiple of magnitude larger, 3. the minimum size required for a database or a CNN, in order to obtain a better result than a random guesser. In this paper, after a solid discussion related to the observed behaviour of CNNs as a function of their sizes and the database size, we confirm that the error's power-law also stands in steganalysis, and this in a border case, i.e. with a medium-size network, on a big, constrained and very diverse database.

show abstract

JPEG Steganalysis Detectors Scalable With Respect to Compression Quality

Abstract: Fast track article for IS&T International Symposium on Electronic Imaging 2020: Media Watermarking, Security, and Forensics proceedings.

Cited by 12 publications

References 0 publications

How to Pretrain for Steganalysis

How to Pretrain for Steganalysis

LSSD: a Controlled Large JPEG Image Database for Deep-Learning-based Steganalysis "into the Wild"

Analysis of the Scalability of a Deep-Learning Network for Steganography "Into the Wild"

Contact Info

Product

Resources

About