We evaluated our previously proposed technique of protecting the copyrights of digital data for 3-D printing. This technique embeds copyright information into not only digital data but also fabricated objects and enables the information to be read to reveal possible copyright violations. The insides of fabricated objects are constructed with fine cavities to embed information into them. In this study, the readability of embedded information relating to the structure parameters of the fine cavities inside fabricated objects was examined to clarify the conditions in which this technique can be applied. The top-view sizes of the cavities, the top-view spaces between cavities, the depths of the cavities themselves, and the depths of the cavities from the surfaces of the objects were changed as experimental parameters. Experimental results clarified the conditions for forming cavities inside the fabricated objects and showed that a sufficient amount of information for copyright (i. e., hundreds of bits) could be embedded if a fabricated object was several centimeters in size.

With the development of cloud computing, data privacy has become a major problem. Reversible data hiding in encrypted images (RDHEI) is an effective technique to embed data in the encrypted domain. Indeed, a lot of methods have been proposed, but none allows a large amount of embedding capacity with a perfect reversibility. In this work, we present a new method of reversible data hiding in encrypted images using MSB (most significant bit) prediction. In order to reconstruct the original image without any errors during the decryption phase, we adapt the to-be-inserted message. Some of the pixels' MSB values are used to highlight the prediction errors and the remaining values are replaced by bits of the secret message. Results show that it is still possible to embed a large message (payload close to 1 bpp).

In this paper, we propose a practical method to estimate the payload rate for individual cover before stego embedding. The proposed method is adopted to the additive distortion model. The a priori knowledge functions employed in the method contains the relation function of steganalyzer's detection error and stego distortion (PE– D), and the relation function of payload rate and distortion(D–α) of the given cover. As it is not suitable to measure the stego security with stego distortion, we adopt PE as the security metric. With the sender's expected PE, the role of PE– D function is to calculate the corresponding D, and then sender can solve out his expectedα with D–α function for the cover. The PE – D function is acquired before estimating phase. During the estimating, the most time-consuming part is calculating the D–α function for the cover, which costs 1 time of stego embedding. Our method is an efficient solution for estimating the secure payload rate.

Typical tasks in a forensic investigation are data acquisition, checksum calculation, file recovery, or content identification. These tasks can be performed mostly without user interaction but are still time-consuming, especially when a large amount of data has to be processed. Individual tasks (or sub-tasks they have in common) often do not perform efficiently and the corresponding implementations could be improved. In this paper we present stream carving, an approach to speed up tasks that are typically performed in a forensic investigation. By identifying and combining similar or identical subtasks and parallelizing most data processing, we are able to decrease the overall processing time significantly. We implemented a stream carving tool that is able to copy, recover, and identify known visual content. The general idea behind stream carving can help developing forensic multi-purpose tools that run several tasks very efficiently.

Partial or selective encryption is a well-known concept in multimedia security. It aims to achieve a level of security of multimedia encryption comparable to common encryption by encrypting only a relevant subset of the complete stream or file. The prime benefit of partial encryption is better performance due to fewer encryption operations. In addition, partially encrypted media data can often be parsed as well as unencrypted media if no header data is encrypted. The focus of partial encryption evaluation has almost always been the level of security that can be achieved. In this work, we discuss another aspect: when partial encryption of MP3 files is used in a DRM scenario, how many resources can be saved by it? As DRM usually is attacked by key sniffing or analogue recording, the security of the encryption itself is of lesser importance as long as it provides a sufficient hindrance to access the media data.

This paper presents a new method, of recompressing a JPEG crypto-compressed image. In this project, we propose a cryptocompression method which allows recompression without any information about the encryption key. The recompression can be executed on the JPEG bitstream directly by removing the last bit of the code representation of a non null DCT coefficient and adapting its Huffman code part. To be able to apply this approach, the quantization table is slightly modified to make up the modifications. This method is efficient to recompress a JPEG cryptocompressed image in terms of ratio compression. Moreover, since the encryption is fully reversible, the decryption of the recompressed image produces an image that has a similar visual quality compared to the original compressed image.

Recent studies have shown that the steganalytic approaches based on deep learning frameworks cannot surpass their rich?model features based companions in peiformance. According to our analysis, one of the main causes of the unsatisfactory performance of deep learning frameworks is that training procedure tends to get stuck at local plateaus or even diverge when starting from a non-ideal initial state. In this paper we will try to investi?gate how to fit deep neural network to a rich-model features set. We regard it as a pre-training procedure and study its 4fect on deep learning for steganalysis. The state-of-the-art JPEG steganalytic features set DCTR is selected as the target and its features extraction procedure is divided into multiple sub-models. A deep learning framework with similar sub-networks is proposed. In the pre-training procedure we train theframeworkfrom bottom to up, fitting the output of each sub-network to the actual output of the corresponding sub-module of DCTR. The motivation behind the scenario is that we reinforce the proposed framework learn to fit the nonlinear mapping implicit in DCTR and expect when it is trainedfrom an initial state which represents an approximate so?lution of DCTR, we can get better peiformance compared to what DCTR has achieved.

Feature-based steganalysis has been an integral tool for detecting the presence of steganography in communication channels for a long time. In this paper, we explore the possibility to utilize powerful optimization algorithms available in convolutional neural network packages to optimize the design of rich features. To this end, we implemented a new layer that simulates the formation of histograms from truncated and quantized noise residuals computed by convolution. Our goal is to show the potential to compactify and further optimize existing features, such as the projection spatial rich model (PSRM).

It is widely recognized that steganography with sideinformation in the form of a precover at the sender enjoys significantly higher empirical security than other embedding schemes. Despite the success of side-informed steganography, current designs are purely heuristic and little has been done to develop the embedding rule from first principles. Building upon the recently proposed MiPOD steganography, in this paper we impose multivariate Gaussian model on acquisition noise and estimate its parameters from the available precover. The embedding is then designed to minimize the KL divergence between cover and stego distributions. In contrast to existing heuristic algorithms that modulate the embedding costs by 1–2| e |, where e is the rounding error, in our model-based approach the sender should modulate the steganographic Fisher information, which is a loose equivalent of embedding costs, by (1–2| e |)^2. Experiments with uncompressed and JPEG images show promise of this theoretically well-founded approach.