Cloud-fog-assisted electronic health record (EHR) systems offer promising solutions for large-scale medical data storage and processing. However, they also raise critical privacy concerns, particularly regarding secure computation over sensitive data, fine-grained bilateral access control, dynamic revocation, and decryption key exposure. Existing cryptographic primitives, such as functional encryption and matchmaking encryption, address some of these challenges individually but fail to offer a unified solution. In this work, we design a revocable and privacy-preserving data computing system with bilateral access control (RPDC-BAC) for cloud-fog-assisted EHR sharing by proposing a novel cryptographic primitive, called server-aided revocable attribute-based matchmaking functional encryption (SR-AB-MFE). Specifically, the proposed scheme supports expressive bilateral access control and computation over encrypted data. In addition, it incorporates time-evolving decryption keys and a server-aided revocation mechanism to mitigate key exposure and efficiently revoke users. To further reduce receiver-side overhead, fog nodes assist in ciphertext authentication and partial decryption. We formally define the proposed primitive and prove its security under static assumptions. Finally, extensive experimental results demonstrate the efficiency and practicality of our design.

Long-term video understanding in intelligent transportation systems (ITS) has advanced significantly with the integration of large language models (LLMs) and vision foundation models. However, existing LLM-based multimodal approaches are limited by context length and memory constraints, restricting their effectiveness to short video scenarios. To address these challenges, we propose AS-Memory, a novel framework that combines adaptive sparse memory with LLMs for efficient and scalable long-term video understanding. AS-Memory introduces a plug-and-play memory bank, a lightweight module designed to seamlessly integrate with existing multimodal LLMs. This memory bank stores and retrieves historical video content, enabling long-term analysis while mitigating context length and GPU memory limitations. To further enhance efficiency, we propose a sparse adaptive mechanism that dynamically compresses redundant features and retains critical information, ensuring effective management of streaming video data. Comprehensive evaluations on long-term video understanding benchmarks demonstrate that AS-Memory consistently outperforms state-of-the-art methods in terms of accuracy. The source code and trained models will be made available to the public.

Vision-Language Models (VLMs) have demonstrated remarkable performance across various tasks by encoding visual frames into tokens analogous to textual tokens, which are then processed by a Large Language Model (LLM) for task execution. To manage computational demands, current methods often employ a token compressor, such as Q-former, for efficient inference. However, these methods are typically trained on video-to-text generation loss, lacking sufficient supervision to align intermediate visual representations with textual semantics, resulting in hallucinations when identifying essential objects. To address this issue, we propose a novel visual-textual alignment framework, Semantic Supervision LLM (SS-LLM), which aligns video and text representations within the intermediate feature space, thereby enhancing the LLM’s decoding process. Additionally, we introduce a CLIP Loss to facilitate visual-text alignment in the intermediate feature space, reducing hallucinations in VLMs. Extensive experiments demonstrate that our approach not only mitigates hallucinations more effectively than existing models but also achieves state-of-the-art performance across several benchmarks, providing more accurate and semantically consistent video-text representations. We will make our source code and trained models publicly available.

Genomic prediction (GP) has made significant progress in the field of breeding. Traditional linear models perform well in handling simple traits but have limitations in extracting nonlinear features for complex traits. The introduction of deep learning (DL) techniques has provided a new approach to GP, especially suited for high-dimensional data processing and complex trait prediction. However, traditional DL models require manual design of the network architecture, which necessitates continuous experimentation and modification. In this paper, we propose a new framework, MDNN, that utilizes the memetic algorithm for neural architecture search and automatically optimizes the network architecture. Compared with the DNNGP, MDNN achieved a 36.49% improvement in the average Pearson correlation coefficient on the wheat599 dataset and a 12.28% improvement on the wheat2000 dataset.

The integration of cloud-edge computing with Internet of Things (IoT) healthcare greatly improves medical service efficiency and reduces home monitoring costs. However, in an untrusted and open environment, it still faces significant privacy and security challenges, especially in terms of confidentiality and authenticity of medical data, as well as bilateral access control between patients and healthcare providers. At present, there are few solutions capable of addressing the aforementioned issues efficiently, as they typically come with significant communication and computational overhead. This poses a substantial challenge for IoT devices that are usually resource-constrained. To address the above issues, this paper proposes an efficient and verifiable fine-grained bilateral access control scheme for cloud-edge IoT healthcare. The scheme adopts flexible attribute-based threshold bilateral access control to provide data confidentiality and authenticity at the same time. In addition, our scheme achieves constant-size ciphertexts, utilizes offline/online technology to accelerate ciphertext generation, and outsources the data authenticity verification and partial decryption process to edge nodes, thereby improving the communication and computational efficiency. Furthermore, our scheme implements verification of outsourced results to resist attacks from malicious edge nodes. The formal security proof and experimental evaluation show that our scheme is more functional and practical than other bilateral access control schemes for IoT healthcare.

Interoperability is a fundamental challenge for longenvisioned blockchain applications. A mainstream approach is using Trusted Execution Environment (TEE) to support interoperable off-chain execution. However, this incurs multiple TEE configured with non-trivial storage capabilities running on fragile concurrent processing environments, rendering current strategies based on TEE far from being practical. This paper aims to fill this gap and design a practical interoperability mechanism with simplified TEE as the underlying architecture. Specifically, we present IvyCross, a TEE-based framework that achieves lowcost, privacy-preserving, and race-free blockchain interoperability. IvyCross allows running arbitrary smart contracts across heterogeneous blockchains atop two distributed TEE-powered hosts. We design an incentive scheme based on smart contracts to stimulate the honest behavior of two hosts, bypassing the requirement of the number of TEE and large memory need. We examine the conditions to guarantee the uniqueness of Nash Equilibrium via Game Theory. Furthermore, an extended optimistic concurrency control protocol is designed to ensure the correctness of concurrent contracts execution. We formally prove the security of IvyCross in the Universal Composability (UC) framework and implement a prototype atop Bitcoin, Ethereum, and FISCO BOCS. Extensive experimental results on end-to-end performance and concurrency control demonstrate the efficiency and practicality of IvyCross.

Federated learning has shown great potential in Internet of Things (IoTs) for performing intelligent decision making. It allows IoT devices to collaboratively train a neural network upon the data they collect while separately keeping these data staying local. However, several research works have shown that such architecture still faces security challenges that adversaries could raise inference attack to the transferring model parameters to reveal data from devices. Moreover, another security risk in federated learning is that malicious devices may launch model pollution attack to reduce the quality of the aggregated model, or dishonest server may output incorrect aggregated result to the devices. Most existing privacy-preserving federated learning protocols could not deal with both problems. In this paper, we present WVFL, a secure weighted aggregation protocol in which aims to minimize the effect of wrong local models to the aggregated model, meanwhile allowing devices to verify the correctness of the aggregation result. All important intermediate values in the process are in encrypted form so that they would not be revealed to both devices and servers to guarantee privacy. At the end of this paper, we give implementation of our WVFL scheme, showing its efficiency compared with previous work.

Blockchain rewriting has become widely explored for addressing data deletion requirements, such as error data deletion, space-saving, and compliance with the “right-to-be-forgotten” rule. However, existing approaches are inadequate for handling cross-chain redaction issues, issues, in facing with the increasing need for inter-chain communication. In particular, transaction rewriting on a blockchain might have relevant effects on the states of other blockchains. The cross-chain interoperability results in inter-chain transactions with more complex dependency relations. The issues pose new challenges to achieve rewriting consistency, for example, ensuring the rewriting of related transactions when a transaction is being modified, and achieve atomic rewriting, whereby two cross-chain transactions must either all, or neither, be processed. This paper introduces a cross-chain solution IvyRedaction, with an emphasis on customizing a decentralized intermediary for generating and maintaining global cross-chain redaction states and transaction dependencies. The paper proposes a novel cross-chain state mapping method with rollback rules, as well as customized block structures and verification algorithms, to address the aforementioned issues. Proof-of-concept experiments are conducted to demonstrate the feasibility of the proposed framework.

The “Right to be Forgotten” rule in machine learning (ML) practice enables some individual data to be deleted from a trained model, as pursued by recently developed machine unlearning techniques. To truly comply with the rule, a natural and necessary step is to verify if the individual data are indeed deleted after unlearning. Yet, previousparameter-spaceverification metrics may be easily evaded by a distrustful model trainer. Thus, Thudiet al. recently present a call to action onalgorithm-levelverification in USENIX Security’22. We respond to the call, by reconsidering the unlearning problem in the scenario of machine learning as a service (MLaaS), and proposing a new definition framework forProof of Unlearning(PoUL) on algorithm level. Specifically, our PoUL definitions (i) enforce correctness properties on both the pre and post phases of unlearning, so as to prevent the state-of-the-art forging attacks; (ii) highlight proper practicality requirements of both the prover and verifier sides with minimal invasiveness to the off-the-shelf service pipeline and computational workloads. Under the definition framework, we subsequently present a trusted hardware-empowered instantiation using SGX enclave, by logically incorporating an authentication layer for tracing the data lineage with a proving layer for supporting the audit of learning. We customize authenticated data structures to support large out-of-enclave storage with simple operation logic, and meanwhile, enable proving complex unlearning logic with affordable memory footprints in the enclave. We finally validate the feasibility of the proposed instantiation with a proof-of-concept implementation and multi-dimensional performance evaluation.

We propose a new approach for privacy-preserving and verifiable convolutional neural network (CNN) testing in a distrustful multi-stakeholder environment. The approach is aimed to enable that a CNN modeldeveloperconvinces auserof the truthful CNN performance over non-public data frommultiple testers, while respecting model and data privacy. To balance the security and efficiency issues, we appropriately integrate three tools with the CNN testing, including collaborative inference, homomorphic encryption (HE) and zero-knowledge succinct non-interactive argument of knowledge (zk-SNARK). We start with strategically partitioning a CNN model into a private part kept locally by the model developer, and a public part outsourced to an outside server. Then, the private part runs over the HE-protected test data sent by a tester, and transmits its outputs to the public part for accomplishing subsequent computations of the CNN testing. Second, the correctness of the above CNN testing is enforced by generating zk-SNARK based proofs, with an emphasis on optimizing proving overhead for two-dimensional (2-D) convolution operations, since the operations dominate the performance bottleneck during generating proofs. We specifically present a new quadratic matrix program (QMP)-based arithmetic circuit witha single multiplication gatefor expressing 2-D convolution operations between multiple filters and inputs in a batch manner. Third, we aggregate multiple proofs with respect to a same CNN model but different testers’ test data (i.e., different statements) into one proof, and ensure that the validity of the aggregated proof implies the validity of the original multiple proofs. Lastly, our experimental results demonstrate that our QMP-based zk-SNARK performs nearly 13.9× faster than the existing quadratic arithmetic program (QAP)-based zk-SNARK in proving time, and 17.6× faster in Setup time, for high-dimension matrix multiplication. Besides, the limitation on handling a bounded number of multiplications of QAP-based zk-SNARK is relieved.

Acting as an important part of Internet of Things (IoTs), vehicular ad-hoc network (VANET) has attracted considerable attention in recent years, where the emergency reporting system is a significant branch and can improve road safety and optimize traffic management. In an emergency reporting system, the authenticity of the emergency messages needs to be ensured carefully since malicious entities may report fake emergency messages to seek personal profit. However, most existing emergency reporting schemes are not efficient enough for secure reporting, given the limited resources of onboard units (OBUs). In this work, we propose a new emergency reporting architecture based on multisignature, where the computation and communication overhead of road-side units (RSUs) is greatly reduced, as the signatures sent to the RSUs have been aggregated. Then, we transform a chameleon signature into a multisignature and propose a secure and OBU-friendly emergency reporting scheme (SOERS) under our architecture. The proposed scheme only requires one hash, two multiplication, and two addition operations to generate a signature for each OBU. Moreover, batch verification of multisignatures further improves the efficiency. Security analysis shows that our scheme is secure against the vehicles-collusion attack and the rogue-key attack, which is a severe issue in multisignature. Finally, performance evaluation shows that our scheme is an efficient solution for emergency reporting.

With the growing popularity of the Internet-of-Things (IoT), a massive amount of purpose-specific, heterogeneous, inexpensive devices have been deployed. To allow these devices to perform their duties and collaborate efficiently, designing a secure and dependable communication channel is necessary. Pairing, as the fundamental procedure for establishing a trustworthy communication channel, has received extensive attention from security researchers. Previous secure pairing protocols depend on auxiliary peripherals (e.g., displays, speakers) to share the secret message, while for those products featuring with low-price, manufacturers would probably adopt insecure pairing methods to reduce the cost, so the devices may be subject to various attacks. To mitigate such a situation, we design a peripheral-free secure pairing protocol, termed SwitchPairing. Our protocol only requires users to connect the pre-pairing devices to the same power source, then randomly presses and releases the switch to generate a shared secret. It does not require additional peripherals and can defense eavesdropping and replay attacks innately. We implement a prototype via two CC2640R2F development boards and invite volunteers to participate in the experiments about bench-marking security and usability. The result of our experiments show that our protocol can fulfill the security and efficient requirement of various IoT applications.

ML-as-a-service (MLaaS) becomes increasingly popular and revolutionizes the lives of people. A natural requirement for MLaaS is, however, to provide highly accurate prediction services. To achieve this, current MLaaS systems integrate and combine multiple well-trained models in their services. Yet, in reality, there is no easy way for MLaaS providers, especially for startups, to collect sufficiently well-trained models from individual developers, due to the lack of incentives. In this article, we aim to fill this gap by building up a model marketplace, called as Golden Grain, to facilitate model sharing, which enforces the fair model-money swapping process between individual developers and MLaaS providers. Specifically, we deploy the swapping process on the blockchain, and further introduce a blockchain-empowered model benchmarking process for transparently determining the model prices according to their authentic performances, so as to motivate the faithful contributions of well-trained models. Especially, to ease the blockchain overhead for model benchmarking, our marketplace carefully offloads the heavy computation and designs a secure off-chain on-chain interaction protocol based on a trusted execution environment (TEE), for ensuring both the integrity and authenticity of benchmarking. We implement a prototype of our Golden Grain on the Ethereum blockchain, and conduct extensive experiments using standard benchmark datasets to demonstrate the practically affordable performance of our design.

Data holders, such as mobile apps, hospitals and banks, are capable of training machine learning (ML) models and enjoy many intelligence services. To benefit more individuals lacking data and models, a convenient approach is needed which enables the trained models from various sources for prediction serving, but it has yet to truly take off considering three issues: (i) incentivizing prediction truthfulness; (ii) boosting prediction accuracy; (iii) protecting model privacy.We design FedServing, a federated prediction serving framework, achieving the three issues. First, we customize an incentive mechanism based on Bayesian game theory which ensures that joining providers at a Bayesian Nash Equilibrium will provide truthful (not meaningless) predictions. Second, working jointly with the incentive mechanism, we employ truth discovery algorithms to aggregate truthful but possibly inaccurate predictions for boosting prediction accuracy. Third, providers can locally deploy their models and their predictions are securely aggregated inside TEEs. Attractively, our design supports popular prediction formats, including top-1 label, ranked labels and posterior probability. Besides, blockchain is employed as a complementary component to enforce exchange fairness. By conducting extensive experiments, we validate the expected properties of our design. We also empirically demonstrate that FedServing reduces the risk of certain membership inference attack.

Android platform is gaining explosive popularity. This leads developers to invest resources to maintain the upward trajectory of the demand. Unfortunately, as the profit potential grows higher, the chances of these Apps getting attacked also get higher. Therefore, developers improved the security of their Apps, which limits attackers ability to compromise upgraded versions of the Apps. However, developers cannot enhance the security of earlier versions that have been released on the Play Store. The earlier versions of the App can be subject to reverse engineering and other attacks. In this paper, we find that attackers can use these earlier versions as attack vectors, which threatens well protected upgraded versions. We show how to attack the upgraded versions of some popular Apps, including Facebook, Sina Weibo and Qihoo360-Cloud-Driven by analyzing the vulnerabilities existing in their earlier versions. We design and implement a tool named DroidSkynet to analyze and find out vulnerable apps from the Play Store. Among 1,500 mainstream Apps collected from the real world, our DroidSkynet indicates the success rate of attacking an App using an earlier version is 34 percent. We also explore possible mitigation solutions to achieve a balance between utility and security of the App update process.

Deep learning can achieve higher accuracy than traditional machine learning algorithms in a variety of machine learning tasks. Recently, privacy-preserving deep learning has drawn tremendous attention from information security community, in which neither training data nor the training model is expected to be exposed. Federated learning is a popular learning mechanism, where multiple parties upload local gradients to a server and the server updates model parameters with the collected gradients. However, there are many security problems neglected in federated learning, for example, the participants may behave incorrectly in gradient collecting or parameter updating, and the server may be malicious as well. In this article, we present a distributed, secure, and fair deep learning framework named DeepChain to solve these problems. DeepChain provides a value-driven incentive mechanism based on Blockchain to force the participants to behave correctly. Meanwhile, DeepChain guarantees data privacy for each participant and provides auditability for the whole training process. We implement a prototype of DeepChain and conduct experiments on a real dataset for different settings, and the results show that our DeepChain is promising.

With cloud computing becoming increasingly popular, there has been a rapid increase in the number of data owners who outsource their data to the cloud while allowing users to retrieve the data. To preserve the privacy of data, data owners usually encrypt their data before outsourcing them to the cloud, and cloud servers can search across the ciphertext domain on behalf of users without learning any information about the data. However, existing work in the literature mostly supports only a single-user or single-keyword search which is not able to satisfy more desired expressive search. Thus, we propose a searchable encryption primitive with attribute-based access control for hybrid boolean keyword search over outsourced encrypted data. There exist several desirable features: (1) Data owners can set search permissions for outsourced encrypted data according to an access control policy. (2) Multiple users, whose attributes satisfy the access control policy, are allowed to perform a retrieval operation upon the encrypted data. (3) Authorized users are able to perform more expressive search, such as any required boolean keyword expression search. Additionally, this primitive is provably secure under our security model and we have also implemented the prototype to show the practicality of the primitive.

Learning deep representations have been applied in action recognition widely. However, there have been a few investigations on how to utilize the structural manifold information among different action videos to enhance the recognition accuracy and efficiency. In this paper, we propose to incorporate the manifold of training samples into deep learning, which is defined as deep manifold learning (DML). The proposed DML framework can be adapted to most existing deep networks to learn more discriminative features for action recognition. When applied to a convolutional neural network, DML embeds the previous convolutional layer's manifold into the next convolutional layer; thus, the discriminative capacity of the next layer can be promoted. We also apply the DML on a restricted Boltzmann machine, which can alleviate the overfitting problem. Experimental results on four standard action databases (i.e., UCF101, HMDB51, KTH, and UCF sports) show that the proposed method outperforms the state-of-the-art methods.