With the rapid development of network media, increasing research on reversible watermarking has focused on improving its robustness to resisting attacks during digital media transmission. There are some other reversible watermarking schemes that work in the encrypted domain for preserving the privacy of the cover image. The robustness of the watermarking and the privacy preserving of the cover image have become the key factors of reversible watermarking. However, there are few robust reversible watermarking schemes in the encrypted domain that could resist common attacks (such as JPEG compression, noise addition) and preserve privacy at the same time. In addition, the embedding capacity of a robust watermark and the efficiency of the encryption method must be considered. Recently, cloud computing technology has led to the rapid growth of network media, and many multimedia properties are owned by multiple parties, such as a film’s producer and multiple distributors. Multi-party watermarking has become an important demand for network media to protect all parties’ rights. In this paper, a Robust Reversible Watermarking scheme in Encrypted Image with Secure Multi-party (RRWEI-SM) based on lightweight cryptography is first proposed. Additive secret sharing and block-level scrambling are developed to generate the encrypted image. Then, the robust reversible watermarking based on significant bit Prediction Error Expansion (PEE) is performed by Secure Multi-party Computation (SMC). For applications with high robustness, a Modified RRWEI-SM is proposed by exploiting two-stage architecture. Furthermore, both the RRWEI-SM scheme and Modified RRWEI-SM scheme are separable and can be applied to multiparty copyright protection. The experimental results and theoretical analysis demonstrate here that the RRWEI-SM and the Modified RRWEI-SM are secure, robust and effective.

Secret Image Sharing (SIS) as a secure data sharing scheme in multiple cover images, has become an increasing researchers' concern. In traditional SIS, the cover image can't be recovered losslessly. The distorted cover images would reduce the practicability of the scheme, especially in medical and military images. The lossless recovery of cover images is required since all details of these images are very critical. In current Invertible SIS (ISIS), the secret image and the cover image may not be reconstructed losslessly. In addition, the authentication capability, visual quality of the stego cover image and embedding rate are limited in spatial domain-based ISIS. As an important carrier, the binary cover image is desired in real applications. Therefore, this paper proposes Transform domain-based Invertible and Lossless Secret Image Sharing schemes with Authentication (T-ILSISA), namely Integer Wavelet Transform-based ILSISA (IWT-ILSISA) and Binarization Transform-based ILSISA (BT-ILSISA) respectively. In (k,n) threshold IWT-ILSISA, the pixels of secret image and the data of cover image are regarded as the coefficients of the (k-1) degree polynomial. The values of generated share are embedded into IWT domain of the cover image. In BT-ILSISA, many different cover images are applied. The generated shares are transformed to the meaningful images since noise-like shares are easy to attract the attacker's attention, are suspected to censors and are difficult for identification and management. In the two schemes, the original secret image and the cover image can be recovered losslessly. The experimental results and theoretical analysis demonstrate that the performances of IWT-ILSISA are better than other similar schemes in the terms of embedding capacity, authentication capability and visual quality of the stego cover image. The BT-ILSISA has a lower computational complexity of the recovery.

The rapid development of cloud computing probably benefits many of us while the privacy risks brought by semi-honest cloud servers have aroused the attention of more and more people and legislatures. In the last two decades, plenty of works seek to outsource various specific tasks to servers while ensuring the security of private data. The tasks to be outsourced are countless; however, the computations involved are similar. In this article, we construct a series of novel protocols that support the secure computation of various functions on numbers (e.g., the basic elementary functions) and matrices (e.g., the calculation of eigenvectors and eigenvalues) on arbitrary $n\geq 2$ servers. All protocols only require constant rounds of interactions and achieve low computation complexity. Moreover, the proposed $n$ -party protocols ensure the security of private data even though $n-1$ servers collude. The convolutional neural network models are utilized as the case studies to verify the protocols. The theoretical analysis and experimental results demonstrate the correctness, efficiency, and security of the proposed protocols.

This paper proposes a high-capacity reversible data hiding scheme for encrypted multimedia data by using Somewhat Homomorphic Encryption. In an image, three adjacent pixels are selected as a group for the whole process. In the encryption part, the original image is encrypted by an image provider. Then, the encrypted image is sent to data hider. In the data hiding part, two absolute differences can be obtained in each group. The additional data are embedded into the encrypted image by shifting histogram of the absolute differences. Moreover, a sorting technique that sorts the sum of two absolute differences in each group has good effects on reducing image distortion. The encryption and embedding operations are controlled by encryption and embedding/data hiding keys, respectively. If a receiver only has the data hiding key, the additional data can be extracted and the ciphertext can be restored. If the receiver has both the decryption key and the data hiding key, the additional data can be extracted and the original image can be restored. The experimental results have demonstrated the superiority of the proposed method over the existing similar methods.