In masked language model (MLM) based attacks, candidate adversarial examples are flexibly generated according to the context, but how to balance imperceptibility and success rate of attacks remains a challenge. To pursue imperceptibility, external semantic constraints are imposed on candidate word generation, whereby the search space is restricted and the success rate of attacks drops. To address this issue, a semantically improved adversarial attack denoted by SAM-CP is proposed by generating high-quality candidate words satisfying the semantic constraints. In particular, linguistic constraints are adopted so that high-quality semantic candidates can be generated as fine-tuning labels instead of manual annotation. Extensive experimental results on three open-source datasets demonstrate that SAM-CP significantly improves the semantic consistency of generated adversarial samples by adopting different MLMs, respectively. Without compromising text quality, the ability to use SAM-CP to deceive state-of-the-art text classifiers is evaluated. Visit https://github.com/HugoTianX/SAM-CP for details and codes.

For security enhancement and privacy preservation, homomorphic encryption is deployed to facilitate computations among cipher texts. To transmit extra data over a cipher text without affecting its usage, lossless data hiding in encrypted domain (LDH-CT) has been developed by exploiting randomness introduced in homomorphic encryption such as Paillier, BGN and NTRU schemes. As data extraction without decrypting the cipher text has been accomplished with BGN, how to retrieve the hidden data after decryption remains unexplored. In this article, an LDH-CT scheme named random number replacement and partitioning (RNRP) is proposed to achieve the versatility of data embedding by designing two algorithms, namely the random number replacement algorithm to embed data to be extracted after decryption and the cipher value selection algorithm to embed data to be extracted without decryption. For the first time, confidential information to be extracted after decryption can be hidden into a cipher text by a third party without knowing its plain text. For convenience in performance evaluation, the proposed algorithms and the RNRP scheme are applied to a set of test images. Experimental results and comparisons with the state-of-the-art schemes demonstrate the better applicability of the proposed scheme, such as in saving bandwidth.

With the proliferation of cloud applications, users are increasingly uploading their private images to cloud servers to avail of supplement storage capacity. Unlike image encryption, the thumbnail-preserving technique is a way of safeguarding image privacy and preserving the outline of the thumbnail image, which allows authorized users to recognize it according to prior knowledge of the original image. Recently, several thumbnail-preserving encryption schemes have been proposed. But most of the sum-preserving encryption-based schemes are irreversible, thereby significantly constraining their practical value. Moreover, the visual quality of images generated with the reversible thumbnail-preserving schemes is unsatisfactory, leading to a decrease in the user recognition accuracy. In light of the aforementioned, this paper presents an efficient scheme to generate thumbnail-preserving images with high visual naturalness, while ensuring reversibility of the process. Experimental results demonstrate that the proposed scheme achieves significant visual quality improvements compared to state-of-the-art schemes. Furthermore, the thumbnail-preserving images generated by using our scheme can effectively evade detection by confusing malicious attackers.

Lossless data hiding in ciphertexts (LDH-CT) is to perform data embedding without changing their plaintexts, which can be used to transmit extra data in the applications of homomorphic encryption at little cost. In this paper, two LDH-CT algorithms named Polynomial Encoding (PE) and Polynomial Modulation (PM) are proposed for the “N-th Degree Truncated Polynomial Ring Unit” (NTRU) scheme, respectively. In the PE algorithm, a polynomial is encoded according to a string of bit values and further used to encrypt a plain-text polynomial. After decrypting the ciphertext, the encoded polynomial can be retrieved so that dozens of bit values can be extracted from it. Moreover, the PE algorithm can be combined with a polynomial partitioning strategy to achieve data extraction before decryption as well. In applying the PM algorithm, no parameter setting of an NTRU cryptosystem is changed while a cipher-text polynomial is generated by selectively sampling a polynomial to match the to-be-hidden value. Furthermore, the data hidden with the PM algorithm can be pre-chosen to be extracted without decryption or after decryption, and in each case up to 10 bit values can be hidden into one cipher-text polynomial. The proposed algorithms and schemes are implemented and compared with several schemes developed for NTRU, BGN, LWE and Paillier encryption. Experimental results and performance evaluations demonstrate the efficacy and superiority of the proposed algorithms and schemes.

Reversible data hiding in ciphertext has potential applications for privacy protection and transmitting extra data in a cloud environment. For instance, an original plain-text image can be recovered from the encrypted image generated after data embedding, while the embedded data can be extracted before or after decryption. However, homomorphic processing can hardly be applied to an encrypted image with hidden data to generate the desired image. This is partly due to that the image content may be changed by preprocessing or/and data embedding. Even if the corresponding plain-text pixel values are kept unchanged by lossless data hiding, the hidden data will be destroyed by outer processing. To address this issue, a lossless data hiding method called random element substitution (RES) is proposed for the Paillier cryptosystem by substituting the to-be-hidden bits for the random element of a cipher value. Moreover, the RES method is combined with another preprocessing-free algorithm to generate two schemes for lossless data hiding in encrypted images. With either scheme, a processed image will be obtained after the encrypted image undergoes processing in the homomorphic encrypted domain. Besides retrieving a part of the hidden data without image decryption, the data hidden with the RES method can be extracted after decryption, even after some processing has been conducted on encrypted images. The experimental results show the efficacy and superior performance of the proposed schemes.

In mobile crowdsensing (MCS), sensing data uploaded by dishonest workers may be false or even malicious. Thus, a reputation management system is often set up by using workers’ historical behaviors to indicate the quality of sensing data. As existing management schemes usually protect the reputation update process, reputation scores are generally stored in plaintext, which may destroy the fair bidding property of an MCS system. To address this issue, we propose an anonymous reputation management system based on the dual blockchain architecture, where reputation scores are masked. More precisely, one chain is used to store and update reputation scores, and another chain is responsible for publishing tasks and storing task-related data. To anonymously update and verify the reputation scores without affecting their usages in data sensing process, a kind of ring signature and Pedersen commitment is employed in smart contracts. In addition, a Schnorr signature is generated to make the reputation scores verifiable in the MCS system. We implement a prototype system on Hyperledger Fabric, and simulation results are provided for comparisons with two existing schemes.

Recently, contrast enhancement with reversible data hiding (CE-RDH) has been proposed for digital images to hide useful data into contrast-enhanced images. In existing schemes, one-dimensional (1D) or two-dimensional (2D) histogram is equalized during the process of CE-RDH so that an original image can be exactly recovered from its contrast-enhanced version. However, noticeable brightness change and color distortion may be introduced by applying these schemes, especially in the case of over enhancement. To preserve image quality, this paper presents a new 2D histogram based CE-RDH scheme by taking brightness preservation (BP) into account. In particular, the row or column of histogram bins with the maximum total height are chosen to be expanded at each time of histogram modification, while the row or column of bins to be expanded next are adaptively chosen according to the change of image brightness. Experimental results on three color image sets demonstrate efficacy and reversibility of the proposed scheme. Compared with the schemes using 1D histogram, image brightness can be preserved more finely by modifying the generated 2D histogram. Moreover, our proposed scheme preserves image color and brightness while achieving better image quality than the existing schemes.

Recently, how to protect privacy in JPEG images has become an important issue in social networks. Besides encryption, reversible visual transformation has been studied for content and privacy protection. In this paper, an improved algorithm is proposed to conceal privacy information in JPEG images. By adopting face detection, the regions to be protected can be identified so that the DC and some AC coefficients in them are modified. The changes that have been made are reversibly saved in the other AC coefficients, and the block selection information is hidden into the whole image also by reversible data hiding. As a secret key is employed to control the process of transformation, the protected content can hardly be obtained without knowing the key. We conduct the proposed algorithm on a set of test images, and compare the performance with several existing algorithms in terms of information leakage, file size increment, and image quality. The experimental results show that improved image quality and less information leakage can be achieved with the proposed reversible transformation algorithm.

Contrast enhancement (CE) of magnetic resonance (MR) brain images is an important technique to bring out the tissue details for clinical diagnosis. Recently, a new form of image enhancement has been proposed to complete the task without any information loss. Specifically, information required to restore the original image is reversibly hidden into the enhanced image. Moreover, several image segmentation based algorithms have been proposed so that the region of interest can be exclusively enhanced. However, with the reversible algorithms, it is hard to properly enhance the tissues in MR brain images when they are relatively small or connected with each other. To address this issue, a hierarchical CE scheme is proposed for MR brain images with reversibility in this letter. Firstly, a deep convolutional neural network is used to segment multiple tissue classes automatically. Then, the segmented tissues are individually utilized to guide the CE procedure so that individual-tissue-enhanced images are generated. Compared with using the background information to guide the CE procedure, better tissue enhancement effects and visual quality are both obtained by our proposed hierarchical scheme. The evaluation results obtained over MR brain test images demonstrate the reversibility and adaptability of the proposed scheme for the enhancement of interested tissues.

With the growth of public demand for online access to health services, many efforts have been devoted to personal health records (PHR) in cloud computing. It enables patients to manage their personal health information (PHI) in cloud servers, which greatly facilitates the collection, access and sharing of PHI. Since cloud servers are not fully trusted, it is desirable that the PHI can be encrypted for privacy protection before uploaded to the cloud. Besides the privacy of PHI, fine-grained and flexible search control is also strongly desired for a secure PHR system. In this article, we first present attribute set based keyword search (ASBKS) which can realize fine-grained keyword search of encrypted PHR. Compared with the existing searchable encryption with access control, the proposed ASBKS can achieve more flexibility in user attributes organization and more efficiency in specifying policies. Furthermore, we present a hierarchical ASBKS scheme to improve scalability by extending ASBKS with a hierarchical structure of users. We implement our ASBKS scheme and the experimental results demonstrate that it is both efficient and flexible for encrypted PHR in cloud computing.

Contrast enhancement (CE) of medical images is helpful to bring out the unclear content in the interested regions. Recently, reversible CE has been proposed so that the original version of a contrast‐changed image can be exactly recovered. This property can be used to save storage space or facilitate the archiving system. To enhance the regions of interest (ROI) without introducing visual distortions, the technique of image segmentation (e.g. using Otsu's method) has been used to obtain the background before conducting the CE process. To segment the ROI more accurately, an interactive algorithm called GrabCut is employed in the proposed scheme. In addition, a new preprocessing strategy is adopted to preserve the image quality through the CE process. Consequently, the content in the selected regions can be better brought out while the reversibility of the CE process is achieved. The experimental results on 30 chest radiograph images and 20 magnetic resonance images have demonstrated the efficacy of the proposed scheme for reversible CE. The evaluation results are provided to show the better performances of the proposed method in achieving CE effects and preserving image quality.

Quantization-based embedding has been used for integrity verification in semi-fragile watermarking. However, some of the illegal modifications cannot be detected in the normal quantization-based methods, especially when the host values (i.e., the values chosen in a host signal for data embedding) are independent from each other. In this paper, a sequential quantization strategy (SQS) is proposed to make the modulation of a host value dependent on a certain number of the previous ones. Therefore, a balance between security improvement and tamper localization can be achieved for integrity verification. Furthermore, the proposed SQS is incorporated with a reversible data hiding mechanism. A new watermarking algorithm is then generated for mesh authentication. The experimental results show that the chance to detect illegal modifications is increased by adopting the SQS while the property of reversibility is achieved.

In this paper, a public-key scheme is proposed to authenticate 3D mesh models. It is well known that digital signature schemes can be used for data authentication by generating a separate signature and appending it to the file. Another way is divide the original data into two parts: the content to be authenticated and the cover content. In the embedding process, the signature of the first part is generated and imperceptibly embedded within the cover content to form the signed content. In the authentication process, a new hash value is produced from the signed content and compared with the value decrypted from the retrieved signature for tamper detection. Before we implement such a scheme on polygonal meshes, mesh partitioning technique is used to divide them into patches with a fixed amount of vertices. For each patch, a corresponding signature is generated to replace the least significant bits of vertex coordinates within it so that the tamper can be localized. The experimental results have shown the promising results

This paper proposes a reversible data hiding method to authenticate 3D meshes by modulating the distances from the mesh faces to the mesh centroid to embed a fragile watermark. It keeps the modulation information in the watermarked mesh so that the reversibility of the embedding process is achieved. Since the embedded watermark is sensitive to geometrical and topological processing, unauthorized modifications on the watermarked mesh can be therefore detected by retrieving and comparing the embedded watermark with the original one. Furthermore, as long as the watermarked mesh is intact, the original mesh can be recovered using some priori knowledge.