Search for: All records

This paper deals with the problem of batch steganography and pooled steganalysis when the sender uses a steganography detector to spread chunks of the payload across a bag of cover images while the Warden uses a possibly different detector for her pooled steganalysis. We investigate how much information can be communicated with increasing bag size at a fixed statistical detectability of Warden’s detector. Specifically, we are interested in the scaling exponent of the secure payload. We approach this problem both theoretically from a statistical model of the soft output of a detector and practically using experiments on real datasets when giving both actors different detectors implemented as convolutional neural networks and a classifier with a rich model. While the effect of the detector mismatch depends on the payload allocation algorithm and the type of mismatch, in general the mismatch decreases the constant of proportionality as well as the exponent. This stays true independently of who has the superior detector. Many trends observed in experiments qualitatively match the theoretical predictions derived within our model. Finally, we summarize our most important findings as lessons for the sender and for the Warden. 

Understanding the mechanisms that lead to false alarms (erro- neously detecting cover images as containing secrets) in steganaly- sis is a topic of utmost importance for practical applications. In this paper, we present evidence that a relatively small number of pixel outliers introduced by the image acquisition process can skew the soft output of a data driven detector to produce a strong false alarm. To verify this hypothesis, for a cover image we estimate a statistical model of the acquisition noise in the developed domain and identify pixels that contribute the most to the associated likelihood ratio test (LRT) for steganography. We call such cover elements LIEs (Locally Infuential Elements). The efect of LIEs on the output of a data-driven detector is demonstrated by turning a strong false alarm into a correctly classifed cover by introducing a relatively small number of “de-embedding” changes at LIEs. Similarly, we show that it is possible to introduce a small number of LIEs into a strong cover to make a data driven detector classify it as stego. Our fndings are supported by experiments on two datasets with three steganographic algorithms and four types of data driven detectors. 

In batch steganography, the sender communicates a secret message by hiding it in a bag of cover objects. The adversary performs the so-called pooled steganalysis in that she inspects the entire bag to detect the presence of secrets. This is typically realized by using a detector trained to de- tect secrets within a single object, applying it to all objects in the bag, and feeding the detector outputs to a pooling function to obtain the final detection statistic. This paper deals with the problem of building the pooler while keep- ing in mind that the Warden will need to be able to de- tect steganography in variable size bags carrying variable payload. We propose a flexible machine learning solution to this challenge in the form of a Transformer Encoder Pooler, which is easily trained to be agnostic to the bag size and payload and offers a better detection accuracy than pre- viously proposed poolers. 

In batch steganography, the sender spreads the secret payload among multiple cover images forming a bag. The question investigated in this paper is how many and what kind of images the sender should select for her bag. We show that by forming bags with a bias towards selecting images that are more difficult to steganalyze, the sender can either lower the probability of being detected or save on bandwidth by sending a smaller bag. These improvements can be quite substantial. Our study begins with theoretical reasoning within a suitably simplified model. The findings are confirmed on experiments with real images and modern steganographic and steganalysis techniques.