Federated learning has recently emerged as a paradigm promising the benefits of harnessing rich data from diverse sources to train high quality models, with salient features that training datasets never leave local devices. Only model updates are locally computed and shared for aggregation produce global model. While federated greatly alleviates privacy concerns opposed centralized data, sharing still poses risks. In this paper, we present system design which offers efficient protection...

Mobile crowdsensing enables convenient sensory data collection from a large number of mobile devices and has found various applications. In the real practice, however, collected are usually unreliable. To extract truthful information unreliable in crowdsensing, topic truth discovery received wide attention recently, which essentially operates by estimating user reliability degrees performing reliability-aware aggregation. Despite effectiveness, applying faces several privacy security...

Truth discovery in mobile crowdsensing has recently received wide attention. It refers to the procedure for estimating unknown user reliability from collected sensory data and inferring truthful information via reliability-aware aggregation. Though widely studied plaintext domain, truth remains largely under-explored privacy-aware crowdsensing. Existing works either do not consider issue or fall short of achieving practical cost efficiency, due iterative transmission computation over large...

Recommendation systems are crucially important for the delivery of personalized services to users. With recommendation services, users can enjoy a variety targeted recommendations such as movies, books, ads, restaurants, and more. In addition, have become extremely effective revenue drivers online business. Despite great benefits, deploying typically requires collection users' personal data processing analytics, which undesirably makes susceptible serious privacy violation issues. Therefore,...

The rapid growth of the Internet Vehicles (IoV) paradigm sparks generation large volumes distributed data at vehicles, which can be harnessed to build models for intelligent applications. Federated learning has recently received wide attentions, allows model training over datasets without requiring raw shared out. However, federated is known vulnerable poisoning attacks, where malicious clients may manipulate local or updates corrupt global model. Such attacks have countered when adopted in...

Cloud computing promises great advantages in handling the exponential data growth. Secure deduplication can greatly improve cloud storage efficiency while protecting confidentiality. In meantime, when are outsourced to remote cloud, there is an imperative need audit integrity. Most existing works only consider support for either secure or integrity auditing. Recently, have been some research efforts aiming integrate with However, prior unsatisfactory that they suffer from leakage of...

Crowdsensing, driven by the proliferation of sensor-rich mobile devices, has emerged as a promising data sensing and aggregation paradigm. Despite useful, traditional crowdsensing systems typically rely on centralized third-party platform for collection processing, which leads to concerns like single point failure lack operation transparency. Such centralization hinders wide adoption wary participants. We therefore explore an alternative design space building atop emerging decentralized...

Neural network (NN) inference services enrich many applications, like image classification, object recognition, facial verification, and more. These NN are increasingly becoming an essential offering from cloud computing providers, where end-users' data offloaded to the for under a customized model. However, current cloud-based operate on clear inputs models, raising paramount privacy concerns. Individual user may contain private information that should always remain confidential. Meanwhile,...

Graphs are widely used to model the complex relationships among entities. As a powerful tool for graph analytics, neural networks (GNNs) have recently gained wide attention due its end-to-end processing capabilities. With proliferation of cloud computing, it is increasingly popular deploy services and resource-intensive training inference in prominent benefits. However, GNN services, if deployed cloud, will raise critical privacy concerns about information-rich proprietary data (and...

Cloud service is a natural choice to store and manage the exponentially produced images. Data privacy one of most concerned points in cloud-based image services. Reversible data hiding over encrypted images (RDH-EI) an effective technique securely confidential cloud. However, existing RDH-EI schemes have obvious weaknesses such as reliable key management system dependence single point failure. To cloud, this study, we propose new reversible strategy via secret sharing. We first design secure...

The explosive growth of multimedia contents, especially videos, is pushing forward the paradigm cloud-based media hosting today. However, wide attacking surface public cloud and growing security awareness from society are both calling for data encryption before outsourcing to cloud. Under circumstance encrypted how still preserve all service benefits center remains be fully explored. In this paper, we present a secure system architecture design as our initial effort toward direction, which...

Public blockchains have emerged as a promising direction in revolutionizing existing data-driven systems relying on centralized service providers. Among others, one kind of such is the popular crowdsensing which promise convenient data collection and aggregation. Although promising, leveraging public to build non-trivial has overcome several barriers. First, are transparent lack support for privacy. Second, participants from open blockchain environment may misbehave serving applications,...

With the rapid advancements in information technology, reversible data hiding over encrypted images (RDH-EI) has become essential for secure image management cloud services. However, existing RDH-EI schemes often suffer from high computational complexity, low embedding rates, and excessive expansion. This paper addresses these challenges by first analyzing block-based secret sharing schemes, revealing significant redundancy within blocks. Based on this observation, we propose two...

Abstract Antiferromagnetic (AFM) spintronics have sparked extensive research interest in the field of information storage due to considerable advantages offered by antiferromagnets, including non-volatile data storage, higher density, and accelerating processing. However, manipulation detection internal AFM order antiferromagnets hinders their applications spintronic devices. Here, we proposed a design idea for an material that is self-assembled from onedimensional (1D) ferromagnetic (FM)...

Two-dimensional carbon-based materials show considerable promise for applications in a wide range of fields, including aerospace, energy storage, and catalysis, due to their great advantages abundant carbon resources, relatively low-cost, non-toxicity, excellent physical chemical properties. However, photovoltaics remain limited. Here, we first theoretically predict stable Sn9C15 monolayer (space group P321). The exhibits numerous advantages, which make it an ideal candidate photovoltaic...

Along with the rapid advancement of digital image processing technology, denoising remains a fundamental task, which aims to recover original from its noisy observation. With explosive growth images on Internet, one recent trend is seek high quality similar patches at cloud databases and harness rich redundancy therein for promising performance. Despite well-understood benefits, such cloud-based paradigm would undesirably raise security privacy issues, especially privacy-sensitive data sets....

Leveraging the wisdom of crowd for knowledge discovery and monetization is increasingly popular nowadays. Among others, one way leveraging crowdsensing with truth discovery, which able to discover truthful from unreliable sensory data harvested mobile clients. In order become truly successful, however, a number challenges are yet be addressed. First, safeguarding clients' demanded privacy protection. Second, in many real applications, usually collected streaming manner, so naturally required...

The wide adoption of cloud greatly facilitates the sharing explosively generated media content today, yet deprives providers' direct control over outsourced content. Thus, it is pivotal to build an encrypted center where only authorized access allowed. Enforcing alone, however, cannot fully protect interests, as users may later become traitors that illegally redistribute public. Such realistic threat should have been seriously treated largely overlooked in literature. In this paper, we...

Large volumes of images are being exponentially generated today, which poses high demands on the services storage, processing, and management. To handle explosive image growth, a natural choice nowadays is cloud computing. However, coming with cloud-based acute data privacy concerns, has to be well addressed. In this paper, we present secure service framework, allows privacy-preserving effective denoising side produce high-quality content, key for assuring quality various image-centric...