Facial expression recognition (FER) remains a challenging task in computer vision. Recent works have shown excellent performance in overall recognition accuracy, but its accuracy significantly decreases when recognizing similar expressions. This is due to interclass homogeneity and intraclass heterogeneity. To address these issues, we propose a novel dual‐stage network called DUAL, inspired by contrastive learning. First, we increase the distance between negative samples while reducing the distance between positive ones. This is achieved by dynamically updating pairs of comparison samples. Second, we introduce a two‐stage network architecture. The first stage uses two branches to extract image features and facial keypoint features. These branches interact to learn coarse‐grained features through mutual guidance. The second stage focuses on fine‐grained features using scale‐specific residual blocks. This allows the model to identify facial regions that are critical for recognizing expressions. We conducted extensive experiments on multiple datasets. The results show that DUAL surpasses state‐of‐the‐art models in items of performance. Additionally, the model shows high accuracy even in noisy conditions, highlighting its robustness.

Location sharing, as an essential component of mobile applications, helps mobile users share location information and enhance their community connections. However, users may be reluctant to share their locations with personal privacy concerns, as anyone including location server who knows these locations can infer much sensitive information about users through analyzing these locations’ information plus their background knowledge. Therefore, it poses a natural question for mobile location-sharing applications how to share users’ locations without any breach of their privacy. To answer this question, in this article, we describe two practical attribute-based encryption schemes served privacy-preserving mobile location-sharing applications. Our proposals are quite suitable for such a mobile application—location sharing with common interests since in our designs users’ interests are also taken into consideration as well as location information. In particular, our two schemes have a higher performance in the sense that in our first construction both ciphertexts and private keys are of constant size simultaneously, and our second construction is an extension of the first that provides a tradeoff between ciphertext size and public-key size. Therefore, the two schemes we designed in this work are quite practical in mobile applications which are often equipped with limited transmission or storage resources. Finally, we offer a formal security proof under a well-defined security model, followed by an experimental evaluation and a theoretical performance comparison.

Visual cryptography (VC) schemes provide a distinguished image encryption technique to protect image security since it can visually decrypt the secret image by superimposing the encrypted shadows. VC schemes for both threshold access structures and general access structures are generally constructed based on the OR operation to minimize the pixel expansion. However, VC schemes with optimal pixel expansion typically have low contrast. Stacking operation OR frequently produces recovered images with poor visual quality and are never able to deliver flawless recovery for secret images. Therefore, we studied XOR-based VC (XVC) schemes that employ linear programming to maximize their contrast. Three schemes for general access structures and three schemes for threshold access structures are designed to maximize their contrast. The proposed schemes’ construction is reduced to a linear programming to maximize the contrast by determining the ideal combinations of basis matrices in terms of primary column matrices and unit matrices, respectively. The comparison study and experimental results demonstrate that the contrast of the previous VC and XVC schemes can be further improved.

Privacy-preserving data evaluation is one of the prominent research topics in the Big Data era. In many data evaluation applications that involve sensitive information, such as the medical records of patients in a medical system, protecting data privacy during the data evaluation process has become an essential requirement. Aiming at solving this problem, numerous fuzzy encryption systems for different similarity metrics have been proposed in literature. Unfortunately, the existing fuzzy encryption systems either fail to achieve attribute-hiding or achieve it, but are impractical. In this article, we propose a new fuzzy encryption scheme for privacy-preserving data evaluation based on overlap distance, which can work in an integer domain while achieving attribute-hiding. In particular, we develop a novel approach to enable an accurate overlap distance to be fast calculated. This technique makes the number of pairing operations during decryption stage negative correlation with the size of the threshold, which is pretty practical for some applications especially with a large threshold. Additionally, we provide a formal security analysis of the proposed scheme, followed by a comprehensive experimental. Also we show that our scheme can be well applied to some scenarios, such as fuzzy keyword searchable encryption and attribute-hiding closest substring encryption.

With the continuous improvement of computation and communication capabilities, the Internet of Things (IoT) plays a vital role in many intelligent applications. Therefore, IoT devices generate a large amount of data every day, which lays a solid foundation for the success of machine learning. However, the strong privacy requirements of the IoT data make its machine learning very difficult. To protect data privacy, many privacy‐preserving machine learning schemes have been proposed. At present, most schemes only aim at specific models and lack general solutions, which is not an ideal solution in engineering practice. In order to meet this challenge, we propose an efficient and privacy‐preserving machine learning training framework (ePMLF) in a fog computing environment. The ePMLF framework can let the software service provider (SSP) perform privacy‐preserving model training with the data on the fog nodes. The security of the data on the fog nodes can be protected and the model parameters can only be obtained by SSP. The proposed secure data normalization method in the framework further improves the accuracy of the training model. Experimental analysis shows that our framework significantly reduces the computation and communication overhead compared with the existing scheme.

A (k, n)-conventional visual cryptography (VC) scheme is designed to share one secret and each participant takes one share. When some common participants are involved in multiple VC schemes for multiple secrets, each needs to take multiple shares. This procedure needs more shares, which is inconvenient. It is desirable that the collaboration between the VC schemes can allow each common participant to keep only one share. Simply merging or gluing together two traditional (k1, n1)and (k2, n2)-VC schemes, after making their pixel expansions the same, might be able to facilitate the collaboration and allow each common participant to keep only one share. But there is a security risk that when a subset of k1participants are from the collection of noncommon participants, some from scheme 1 and some from scheme 2, they can reconstruct secret 1, which is inconsistent with the intention of the original scheme. Similarly, k2noncommon participants could reconstruct secret 2. This shortcoming is inherited from the brute-force combination of traditional schemes. Therefore, a more sophisticated mechanism is required; this is the main task of this paper. In this paper, we first transform collaborative VC (CVC) schemes into the multiple secrets VC scheme with a general access structure. The construction of the basis matrices in CVC scheme between two VC schemes is formulated into an integer linear programming problem that minimizes the pixel expansion under the corresponding security and contrast constraints. Also the collaboration among more VC schemes is constructed. Finally, the experimental results illustrate the construction procedure of the CVC scheme and demonstrate the effectiveness of the CVC scheme.