Search

With the widespread use of cloud emails and frequent reports on large-scale email leakage events, a security property so-called forward secrecy becomes desirable and indispensable for both individuals and cloud email service providers to strengthen the security of cloud email systems. However, due to the failure to meet the security and practicality requirements of email systems simultaneously, typical methods of achieving forward secrecy, such as Diffie-Hellman key exchange and forward-secure public-key encryption, have not been widely approved and adopted. In this paper, to capture forward secrecy of encrypted cloud email systems without sacrificing the practicability, we introduce a new cryptographic primitive named forward-secure puncturable identity-based encryption (fs-PIBE), which enables an email user to perform fine-grained revocation of decryption capacity. Based on this primitive, we build a framework of encrypted cloud email systems, and instantiate it with a concrete fs-PIBE construction that has constant size of ciphertext and provable security in the standard model. Furthermore, to improve the security and efficiency of the presented framework, we extend the proposed scheme to support end-to-end encryption and outsourced decryption, respectively. In addition, as a proof-of-concept of the proposed fs-PIBE scheme, we implement it and produce various experiments to demonstrate its practicability and correctness.

Bronchial asthma is a common clinical disease in pediatrics. It affects brain function and causes cognitive impairment due to long-term hypoxia in the brain. Studies have shown attention deficits in asthma children, but lack the basis for brain studies. Functional magnetic resonance imaging (fMRI) based on functional connectivity analysis was used in this study, and 31 asthma children and typically developing children (TDC) were compared in degree centricity (DC), voxel-mirrored homotopic connectivity (VMHC) and continuous performance test (CPT). The study found that the correct number of CPT in the asthma group was significantly lower than the TDC group. And the reaction time in the asthma group was significantly longer than the TDC group. Compared with TDC group, asthma children exhibited lower DC value in the right superior frontal gyrus (after FDR correction, P

Encryption and compression are two critical techniques to ensure data confidentiality and efficiency for a cloud-based data storage system, respectively. However, directly combing encryption and compression incurs substantial performance degradation. We propose TinyEnc, an encrypted data storage system for cloud-based key-value store. TinyEnc supports encryption and compression simultaneously with rich query support.To reconcile encryption and compression without compromising performance, we propose a new encrypted compression data structure to enable fine-grained access to compressed and encrypted key-value data. We then propose two new transforming mechanisms, namely orthogonal data dividing and hierarchical data padding, to transform a plaintext key-value table into the encrypted compression data structure in a privacy-preserving way. Finally, we craft order-revealing encryption (ORE) and symmetric searchable encryption (SSE) to design a new encrypted search index over the encrypted compression data structure to support rich types of data queries. We implement a prototype of TinyEnc on top of Cassandra. Our evaluation shows that TinyEnc increases the throughput by up to 7 times and compression ratio by up to 1.3 times with respect to previous works.

Symmetric Searchable Encryption (SSE) schemes facilitate searching over encrypted data, and have been extensively explored to improve function, efficiency or security. There are, however, additional functions that we need to consider in a real-world setting. For example, forward and backward privacy are required to adequately secure newly added documents and deleted documents in Dynamic SSE (DSSE) schemes, and support boolean search (that allows users to search over encrypted data using basic boolean operations) to achieve improved efficiency and retrieval accuracy. Therefore, in this paper we first construct the Verifiable Boolean Search over encrypted data (VBS), and then improve VBS to achieve Forward and Backward privacy (VBS-FB). Finally, we formally prove the security of our proposed schemes, and evaluate their performance using real-world datasets.

Searchable Symmetric Encryption (SSE), which enables users to search over encrypted data without decryption, has gained increasing attention from both academic and industrial fields. However, existing SSE schemes either have low search efficiency or cannot support multi-keyword search, dynamic updates, and result verification simultaneously. To solve these problems, we propose a Verifiable and Dynamic Multi-keyword Search (VDMS) scheme over encrypted data by using the bitmap and RSA accumulator, which provides multi-keyword search over encrypted data in an efficient, verifiable and updated way. The bitmap is used as a data structure to build the indexes, which improves the search efficiency and reduces the storage space of the indexes. The RSA accumulator and bitmap are combined to verify the correctness of results. Formal security analysis proves that our VDMS is adaptively secure against Chosen-Keyword Attacks (CKA), and empirical experiments using a real-world dataset demonstrate that our VDMS is efficient and feasible in practical applications.

Blockchain allows for secure management of a shared ledger by agreement protocols, where transactions are validated over network without central authorities. Although the agreement protocol has been thoroughly conducted of propagation and consensus researches, mobility of nodes in unstructured overlay networks has not received much attention. Besides, current dynamic propagation schemes waste travel hops and are low of delivery ratio. In this paper, we propose a social characteristic-based propagation-efficient protocol NefSBFT to agree on system state plus consensus mechanisms in public blockchains. We devise a propagation technique (travel hops of at least $\frac{1}{3}$ savings, delivery ratio above $0.93$ , etc.) for message multicasting when exploiting real nodes' social characteristics of intermittent connectivity and frequent partitions. This propagation technique is executed in the improved FastBFT to achieve transaction ordering and block verification, thus, no controllable mobility is required during the whole system's execution. NefSBFT achieves fast propagation, small message complexity and few resource consumption of travel hops and running nodes for complete protocol execution. We analyze NefSBFT's security against DDOS attack of non-primary failure. The experiments show the performance tradeoff under different parameters, compare the propagation efficiency with Erlay and Flooding, and clarify NefSBFT's impact on the whole system performance through comparison.

Ranked keyword search has gained increasing attention due to its attractive advantages such as flexibility and accessibility. However, most existing ranked keyword search schemes ignore the semantic associations between the documents and queries. To solve this challenging issue in cloud-assisted edge computing, we first design the Semantic-aware Ranked Multi-keyword \textbf{S}earch (SRMS) scheme by adopting the Latent Dirichlet Allocation (LDA) topic model and the Chinese Remainder Theorem (CRT)-based secret sharing mechanism. Considering that the cloud server may be malicious, we implement a basic verification mechanism in SRMS to verify the correctness and completeness of search results and extend this verification mechanism in cloud-assisted edge computing scenarios. Formal security analysis proves that SRMS and extended result verification mechanisms are secure in both the known ciphertext model and the known background model. Extensive experiments using the real-world dataset demonstrate that SRMS is efficient and feasible in practical applications.

Worker selection in crowdsensing plays an important role in the quality control of sensing services. The majority of existing studies on worker selection {\color{red}were} largely dependent on a trusted centralized server, which might suffer from single point of failure, the lack of transparency and so on. Some works recently proposed blockchain-based crowdsensing, which {\color{red}utilized} reputation values stored on blockchains to select trusted workers. However, the transparency of blockchains enables attackers to effectively infer private information about workers by the disclosure of their reputation values. In this paper, we {\color{red}proposed} the TrustWorker, a trustworthy and privacy-preserving worker selection scheme for blockchain-based crowdsensing. By taking the advantages of blockchains such as decentralization, transparency and immutability, our TrustWorker {\color{red}could} make the worker selection process trustworthy. To protect workers' reputation privacy in our TrustWorker, we {\color{red}adopted} a deterministic encryption algorithm to encrypt reputation values and then {\color{red}selected} the top N workers in the light of secret minimum heapsort scheme. Finally, we theoretically {\color{red}analyzed} the effectiveness and efficiency of our TrustWorker, and then {\color{red}conducted} a series of experiments. The theoretical analysis and experiment results demonstrate that our TrustWorker can achieve trustworthy worker selection, while ensuring the workers' privacy and the high quality of sensing services.

Accurate medical image segmentation of brain tumors is necessary for the diagnosing, monitoring, and treating disease. In recent years, with the gradual emergence of multi-sequence magnetic resonance imaging (MRI), multi-modal MRI diagnosis has played an increasingly important role in the early diagnosis of brain tumors by providing complementary information for a given lesion. Different MRI modalities vary significantly in context, as well as in coarse and fine information. As the manual identification of brain tumors is very complicated, it usually requires the lengthy consultation of multiple experts. The automatic segmentation of brain tumors from MRI images can thus greatly reduce the workload of doctors and buy more time for treating patients. In this paper, we propose a multi-modal brain tumor segmentation framework that adopts the hybrid fusion of modality-specific features using a self-supervised learning strategy. The algorithm is based on a fully convolutional neural network. Firstly, we propose a multi-input architecture that learns independent features from multi-modal data, and can be adapted to different numbers of multi-modal inputs. Compared with single-modal multi-channel networks, our model provides a better feature extractor for segmentation tasks, which learns cross-modal information from multi-modal data. Secondly, we propose a new feature fusion scheme, named hybrid attentional fusion. This scheme enables the network to learn the hybrid representation of multiple features and capture the correlation information between them through an attention mechanism. Unlike popular methods, such as feature map concatenation, this scheme focuses on the complementarity between multi-modal data, which can significantly improve the segmentation results of specific regions. Thirdly, we propose a self-supervised learning strategy for brain tumor segmentation tasks. Our experimental results demonstrate the effectiveness of the proposed model against other state-of-the-art multi-modal medical segmentation methods.

With the advent of cloud computing, explosively increasing images are gradually outsourced to the cloud server for costs saving and feasibility. For security and privacy concerns, images (e.g., medical diagnosis, personal photos) should be encrypted before being outsourced. However, traditional encrypted image retrieval techniques still suffer from costly access control and low search accuracy. To solve these challenging issues, in this paper, we first propose a Controllable encrypted cloud image Search scheme in Multi-user settings (namely CSM) by using the polynomial-based access strategy and proxy re-encryption technique. CSM achieves efficient access control and avoids heavy communication overhead caused by key transmission. Then, we improve the basic CSM to achieve malicious search user Tracing (namely TCSM) by utilizing the watermark technique, which can further prevent search users from illegally redistributing retrieved images to unauthorized search users. Our formal security analysis proves that our CSM (or TCSM) can guarantee the privacy of images, indexes, and search queries. Our empirical experiments using real-world datasets demonstrate the efficiency and high accuracy of our CSM (or TCSM) in practice.

Coarse-grained reconfigurable architecture (CGRA) is a promising programmable hardware with high power-efficiency and high performance. However, compiling and optimizing loops with irregular branches on CGRAs is a challenge to fulfill the performance potential. Existing predication techniques, such as partial predication (PP) and full predication (FP), conservatively implement software pipeline with a static initiation interval (II) obtained from the maximum graph, and thus only parts of the graph in each loop iteration will be actually executed, resulting in underexploited performance. To exploit more loop-level parallelism for irregular branches, this paper proposes a novel dynamic-II pipeline (DIP) scheme, which realizes a pipeline with variable II by accommodating multiple iterations of short path in one static configuration. Since the DIP scheme is effective to only certain types of branches, this paper designs a hybrid compilation framework integrating other complementary methods, which selects appropriate method for source programs according to a proposed performance evaluation model. Experimental results show that 1) the hybrid compilation framework can effectively extract branch features, correctly choose and implement corresponding branch processing methods within acceptable compile time, 2) as compared to PP and FP, DIP brings a significant total execution time (TET) reduction by 27.21% and 22.04% on average when execution probability of short branch is 50%.

With the explosive growth of data volume in the cloud computing environment, data owners are increasingly inclined to store their data on the cloud. Although data outsourcing reduces computation and storage costs for them, it inevitably brings new security and privacy concerns, as the data owners lose direct control of sensitive data. Meanwhile, most of the existing ranked keyword search schemes mainly focus on enriching search efficiency or functionality, but lack of providing efficient access control and formal security analysis simultaneously. To address these limitations, in this paper we propose an efficient and privacy-preserving Multi-keyword Ranked Search scheme with Fine-grained access control (MRSF). MRSF can realize highly accurate ciphertext retrieval by combining coordinate matching with Term Frequency-Inverse Document Frequency (TF-IDF) and improving the secure kNN method. Besides, it can effectively refine users' search privileges by utilizing the polynomial-based access strategy. Formal security analysis shows that MRSF is secure in terms of confidentiality of outsourced data and the privacy of index and tokens. Extensive experiments further show that, compared with existing schemes, MRSF achieves higher search accuracy and more functionalities efficiently.

Fog-assisted Internet of Things (IoT) can outsource the massive data of resource-constraint IoT devices to cloud and fog nodes. Meanwhile, it enables convenient and low time-delay data sharing services which relies heavily on high security of data confidentiality and fine-grained access control. Many efforts have been focused on this urgent requirement by leveraging Ciphertext-Policy Attribute-Based Encryption (CP-ABE). However, when deployed in Fog-assisted IoT systems for secure data sharing, it remains a challenging problem that how to preserve attribute privacy of access policy, and trace-then-revoke traitors (i.e., malicious users intending to leak decryption keys for illegal profits) efficiently and securely in such a large scale and decentralized environment with resource-constraint user devices, especially in consideration of misbehaving cloud and fog nodes. Therefore, in this paper, we propose a revocable and privacy-preserving decentralized data sharing framework (RPDDSF) by designing a large universe and multi-authority CP-ABE scheme with fully hidden access policy for secure data sharing in IoT systems to achieve user attribute privacy preserving with unbounded attribute universe and key escrow resistance suitable for large scale and decentralized environment. Based on this, with RPDDSF, anyone can efficiently expose the traitors and punish them by forward/backward secure revocation. Besides, RPDDSF is able to guarantee data integrity for both data owners and users to resist misbehaving cloud and fog nodes, alongwith low computation overhead for resource-constraint devices. Finally, RPDDSF is proven to be secure with detailed security proofs, and its high efficiency and feasibility are demonstrated by extensive performance evaluations.

Wireless networks ensure the Unmanned Aerial Vehicles (UAVs) communicate and cooperate with each other, which plays an indispensable role among UAVs. The two crucial challenges in UAV wireless networks are energy saving and security. The current light-weight communication approaches lead to insufficient robustness of the encrypted transmission that is insecure. To address this issue, we propose a secure transmission approach with energy-efficiency towards UAVs networks. We design a light-weight symmetric encryption algorithm based on SM4 and the relevant key negotiation and update mechanism to protect the confidentiality of communication contents. Moreover, a modified aggregative BLS signature scheme, together with Merkle Hash Tree (MHT) is introduced to guarantee the integrity and authenticity of data packets in transmission. Furthermore, we propose an online/offline revocable identity-based group signature (OORIBGS) scheme and integrate it into our framework for UAV anonymity, traceability as well as revocability with small key management cost and high efficiency. We give detailed security analysis and prove that our proposal has the properties of data confidentiality, integrity and authenticity as well as identity traceability and anonymity. Moreover, we apply our approach in the UAVs networks and evaluate the runtime and anti-attack performance. The experimental results show that the proposed method can be effectively used in UAVs secure communication.

In cloud-assisted electronic health (eHealth) systems, outsourced electronic health records (EHRs) have attribute-value type formats: an EHR corresponds to an entry with a unique identity and has multiple types of attribute values. Such formatted EHRs form an attribute-value type database, where both the attribute values and unique identities can serve as keywords for searching. Since EHRs are very sensitive, they are always encrypted before being outsourced, which makes retrieval of target EHRs by either identity or attribute value hard. Moreover, the EHRs are encrypted by different doctors, a researcher, who is delegated to research a certain kind of disease, cannot find out all corresponding EHRs due to the difference of encryption keys. In this article, we construct a triple dictionary index structure for the attribute-value type database to allow a researcher to retrieve encrypted EHRs by the identity and attribute value and to perform dynamic operations over them. We employ an identity server to assist doctors in generating encryption keys via an oblivious way. By doing so, the researcher can retrieve encrypted EHRs without leaking any information to the identity server. We analyze the security and evaluate the performance of our scheme to demonstrate that it achieves $L$ -adaptive security and forward security with high efficiency.

Edge computing is a new computing paradigm, whichprovides computing resources at the edge of the network close to the data source and reduces the response delay and data transmission bandwidth. Data aggregation is a fundamental computing service in edge computing. However, existing data aggregation schemes do not consider the problem of fine-grained data aggregation. Besides, most schemes use a pair-based signature scheme, which cannot meet the actual situation of limited resources of terminal devices. In this article, we propose a lightweight, fine-grained privacy-preserving data aggregation scheme. The terminal devices add the user’s characteristic identifier bit in the generated message. Therefore, the edge servers can combine the characteristic identifiers in the messages to perform fine-grained data aggregation according to the aggregation rule list to obtain aggregated data of multiple target groups. We also use the unpaired signature scheme based on the elliptic curve, which has a faster signature and verification speed. Also, we use masks to avoid the privacy leakage of user’s characteristic identifiers. According to the experiments, the computation and communication overhead of our scheme is low. Simultaneously, the cloud center can dynamically update the aggregation rule list sent to the edge server according to the demands.

A finite element simulation process based on methods of mutual calling between MATLAB and ANSYS software packages are proposed, which has the ability to consider the influence of tension insulator string weight on the sag of a span in a tension section. The convergence criterion of the process can be controlled artificially and is defines to converge when the maximum sag of two calculation steps is very small. A simplified theoretical analysis method is also developed for comparison and verification. These proposed methods are then applied to a tension section and a three-span actual line. Results show that when using a finite element simulation to obtain the curve of a wire, if the number of elements in the span is 10, the mid-span sag in the span is consistent with the theoretical result. And if the number of elements in the span is greater than 50, the sag curve of the conductor is almost unchanged. The two methods and the difference between them are very small, indicating the efficiency of the proposed process.

Collaborative learning has been used to train a joint model on geographically diverse data through periodically sharing knowledge. Although participants keep the data locally in collaborative learning, the adversary can still launch inference attacks through participants’ shared information. In this paper, we focus on the property inference attack during model training and design a novel defense mechanism, namely NOSnoop, to defend such an attack. We propose a collaborative meta-learning architecture to learn the common knowledge over all participants and utilize the natural advantage of meta-learning to hide the sensitive property data. We consider both irrelevant property and relevant property preservation in NOSnoop. For irrelevant property preservation, we utilize the inherent advantage of meta-learning to hide the sensitive property data in meta-training support dataset. Thus, the adversary cannot capture the key information related to the sensitive properties and cannot infer victim’s private property successfully. For relevant property preservation, an adversarial game is further proposed to reduce the inference success rate of the adversary. We conduct comprehensive experiments to evaluate the effectiveness of NOSnoop. When hiding the sensitive property data in meta-training support dataset, NOSnoop achieves an inference AUC score as low as 0.4984 for irrelevant property preservation, meaning the adversary cannot distinguish whether the training batch has the sensitive property data or not. When preserving the relevant property, NOSnoop is able to achieve an inference AUC score of 0.5091 without compromising model utility.

Vehicular networks have tremendous potential to improve the road safety and traffic efficiency, and the adoption of the Space-Air-Ground Integrated Network (SAGIN) architecture in vehicular networks can greatly improve the performance of vehicular networks by leveraging the respective advantages of the space, air, and ground segments on coverage, flexibility, reliability, and availability, which results in Space-Air-Ground Integrated Vehicular Networks (SAGIVNs). Trust management is an important tool for constructing trustworthy SAGIVNs, and privacy preservation is also a primary concern in SAGIVNs. They have conflicting requirements and a satisfactory balance between them is urgently required. In this article, we propose a novel Privacy-Preserving Trust Management (PPTM) scheme for the emergency message dissemination in SAGIVNs. The proposed scheme can realize precise trust management and strong conditional privacy preservation simultaneously with low communication overhead, and can provide strong applicability, strong robustness, and multiple other attractive features. Furthermore, the exhaustive theoretical analysis and simulation evaluation are detailed. The results reveal that the proposed scheme is significantly superior to the existing schemes in several aspects.

Federated Learning (FL), a cooperative distributed learning framework, has been employed in various intelligent Internet-of-Things (IoT) applications (e.g., smart health-care, smart home, smart industry). However, there may be malicious devices in these IoT applications inferring other devices’ privacy or destroying the uploaded model parameters. Besides, due to the heterogeneity of IoT devices, it is difficult for the existing synchronized FL to effectively train models through non-identical independently distributed (non-IID) local datasets. To address these issues, we propose an asynchronous grouped federated learning framework (PAG-FL) for IoT, enabling multiple devices and the server to collaboratively and efficiently train models without revealing privacy. PAG-FL framework consists of an adaptive Renyi Differential Privacy based privacy budget allocation (ARB) protocol and an asynchronous weight-based grouped update (AWGU) algorithm. In particular, our ARB protocol applies Renyi Differential Privacy and adaptively adjusts the privacy budget to obtain an efficient local model. The AWGU algorithm can defend against the poisoning attack on non-IID dataset by weighing grouped local models to generate a global model. Meanwhile, it also realizes the asynchronous optimized update by adopting a lazy loading strategy. Theoretically, the proposed framework has a convergence guarantee and a privacy guarantee when training over the non-IID dataset in an asynchronous FL. Our empirical experiments validate the effectiveness of the theoretical design and demonstrate the improved utility and robustness of PAG-FL in heterogeneous IoT.

With the emergence of intelligent terminals, the Content-Based Image Retrieval (CBIR) technique has attracted much attention from many areas (i.e., cloud computing, social networking services, etc.). Although existing privacy-preserving CBIR schemes can guarantee image privacy while supporting image retrieval, these schemes still have inherent defects (i.e., low search accuracy, low search efficiency, key leakage, etc.). To address these challenging issues, in this paper we provide a similarity Search for Encrypted Images in secure cloud computing (called SEI). First, the feature descriptors extracted by the Convolutional Neural Network (CNN) model are used to improve search accuracy. Next, an encrypted hierarchical index tree by using K-means clustering based on Affinity Propagation (AP) clustering is devised, which can improve search efficiency. Then, a limited key-leakage k-Nearest Neighbor (kNN) algorithm is proposed to protect key from being completely leaked to untrusted image users. Finally, SEI is extended to further prevent image users' search information from being exposed to the cloud server. Our formal security analysis proves that SEI can protect image privacy as well as key privacy. Our empirical experiments using a real-world dataset illustrate the higher search accuracy and efficiency of SEI.

The rapid development of cloud computing and Internet of Things enables widely use of Internet of Vehicles (IoV) and drives a revolutionary change for intelligent transport system. The advent of 5G and fog computing architectures also facilitates applications of IoV and makes the enormous IoV data outsourcing and sharing in cloud and fog more convenient. However, the data outsourcing and sharing will incur many security concerns to IoV including data leakage and breach as the users cannot direct control their data. To guarantee IoV data security, a novel scheme is proposed for IoV data sharing with high efficiency. In our scheme, we utilize ciphertext-policy attribute-based encryption (CP-ABE) to achieve confidentiality and fine-grained access control for IoV data shared in cloud and fog. Nevertheless, as another intrinsic feature in IoV environment, the dynamicity in a group of connected vehicles brings some challenges to our scheme. To deal with this situation, we bring forward a concept of auditable user revocation for CP-ABE to adapt to the dynamic vehicle groups. Moreover, online/offline and verifiable outsourcing techniques are leveraged to improve the efficiency and guarantee the correctness of decryption. At the end, we demonstrate the security and efficiency for our scheme by giving detailed security analysis and performance evaluation with extensive experiments.

Edge services provide an effective and superior means of real-time transmissions and rapid processing of information in the Industrial Internet of Things (IIoT). However, the continuous increase in the number of smart devices results in privacy leakage and insufficient model accuracy of edge services. To tackle these challenges, in this research, we propose a blockchain-based machine learning framework for edge services (BML-ES) in IIoT. Specifically, we construct novel smart contracts to encourage multi-party participation of edge services to improve the efficiency of data processing. Moreover, we propose an aggregation strategy to verify and aggregate model parameters to ensure the accuracy of the decision tree learning model. Finally, based on SM2 public-key cryptosystem, we protect data security and prevent data privacy leakage in edge services. Theoretical analysis and simulation experiments indicate that BML-ES framework is effective and efficient, and is better suitable to improve the accuracy of edge services in IIoT.

Searchable Encryption (SE) enables a data owner to outsource encrypted data to an untrusted server while preserving keyword search functionality. Typically, the server learns whether or not a query has been performed more than once, which is called the search pattern. However, such kind of information leakage might be leveraged to break query privacy. To reduce such type of leakage and provide strong privacy guarantee, Wang et al. proposed a novel SE scheme based on the Paillier encryption scheme in INFOCOM'15. Unfortunately, their scheme cannot perform keyword search successfully, because the additive homomorphic property is not sufficient for their construction. In this paper, we first show that why their scheme fails to return the correct search result, and then propose a new SE scheme by adopting a special additive homomorphic encryption scheme to achieve the multiplicative homomorphic property efficiently. Furthermore, we enhance the security on the user side. Specifically, we use random polynomials with an appropriate degree to guarantee that the user cannot learn anything other than the desired search result. Finally, we present a formal security analysis and implement our scheme on a real-world database, which demonstrates that our construction can achieve the desired security properties with good performance.

Controller Area Network (CAN) bus anomaly detection on the Internet of Vehicle (IoV) is a topic of ongoing interest and increasing importance, particularly as IoV becomes commonplace. However, existing abnormal detection schemes are not ideal due to the lack of abnormal data in IoV and the difficulty of parsing the diverse message rules in existing vehicles. While single-classification of the SVDD (Support Vector Domain Description) algorithm can detect abnormalities only with normal message information, the approach has a high rate of false negatives when deployed directly in the in-vehicle network environment. In addition, the real-time data generation (e.g., IoV messages) compounds the challenge of designing effective detection approaches. Therefore, this paper proposes a mechanism for car networking message classification for a broad range of data, and establishes a weak model for many simple redundant data in the vehicle intranet. The detection can reduce the time and computation costs of detection while ensuring accuracy. Then, this paper proposes two improved SVDD schemes: M-SVDD scheme which adds the Markov chain to detect time-related messages; G-SVDD scheme which maps the kernel function to the Gaussian kernel function to reduce the redundant area of the model and the false-negative rate. The experimental results show that our proposed schemes have higher accuracy, recall rate and fewer computation overhead.

Nowadays, hardware security has become a serious concern for modern CPUs. State-of-the-art detection approaches rely heavily on trustworthy and intimate internal states, incurring significant design/operation overheads and additional risks to security and intellectual property. This article proposes an architecture called Jintide, which utilizes trusted external monitors to validate an untrusted CPU chip at runtime. This architecture records, replays, and analyzes the CPU’s IO and memory behavior with the architectural states. The Jintide simultaneously verifies whether the records are correctly replayed with the instruction set architecture and whether the records involve malicious behavior. Consequently, not only architectural but also micro-architectural threats can be detected. The Jintide adopts the states from the untrusted source because it has a built-in function to detect spurious states. The monitors comprise three types of chips (with 28-/40-nm TSMC technology): a tracer chip to record the behavior of IO ports, multiple tracer chips to record the behavior of DDR4 DIMMs, and a reconfigurable chip to verify these records with software states. As runtime external monitors, the Jintide would be especially suitable to constitute distributed large-scale clusters, which can amortize operation overheads. This scheme is effective in detecting pervasive hardware security issues, including vulnerabilities, backdoors, and hardware Trojans. The measured results show that a system composed of 300 000 Jintide CPUs containing Intel Xeon Skylake processors can detect over 99.8% of recognizable attacks at the cost of 0.98% performance loss. Hence, the Jintide is an extensible, low-cost, and effective solution to improve the hardware security of large-scale CPU clusters.

Convolutional neural networks (CNNs) have become the critical technology to realize face detection and face recognition in the face tracking (FT) system. However, traditional CNNs usually have nontrivial computational time and high energy consumption, making them inappropriate to be deployed in the large-scale time-sensitive FT system. To address this challenge, we design an artificial intelligence and Internet of Things (AIoT) empowered edge-cloud collaborative computing (ECCC) system based on the energy-efficient field-programmable gate array (FPGA)-based CNN accelerators for the purpose of realizing a low-latency and low-power FT. First, we present the AIoT-empowered ECCC system architecture, which consists of an intelligent computing subsystem, an Internet-of-Things (IoT) subsystem, an edge-cloud collaborative subsystem, and an application subsystem. In what follows, we investigate the enabling technologies for these subsystems. Thereafter, we develop an FPGA-based hardware accelerator dedicated to the compact MobileNet CNN by using the hardware design techniques, such as systolic array, matrix tiling, fixed-point precision, and parallelism. Furthermore, we integrate the FPGA accelerators with CPUs and GPUs to build a context-aware CPU/GPU/FPGA heterogeneous computing system. Finally, we implement a delay-aware energy-efficient scheduling algorithm dedicated to this heterogeneous system. With the above hardware and software codesign mechanism, the energy cost and execution time of CNNs can be decreased significantly. The real-world experiments on the CPU/GPU/FPGA-based ECCC system proved the effectiveness of the proposed schemes in reducing the latency and improving the power efficiency of the FT system.

With the large-scaled data generated from various interconnected machines and networks, Industrial Internet of Things (IIoT) provides unprecedented opportunities for facilitating data mining for industrial applications. The current IIoT architecture tends to adopt cloud computing for further timely mining IIoT data, however, the openness of security-critical IIoT becomes challenging in terms of unbearable privacy issues. Most existing privacy-preserving data mining (PPDM) techniques are designed to resist honest-but-curious adversaries (i.e., cloud servers and data users). Due to the complexity and openness in IIoT, PPDM is significantly difficult with the presence of malicious adversaries in IIoT who may incur incorrect learned models and inference results. To solve the aforementioned issues, we propose a framework to extend existing PPDM to guard linear regression against malicious behaviors (hereafter referred to as GuardLR). To prevent dishonest computations of cloud servers and inconsistent inputs of data users, we first design a privacy-preserving verifiable learning scheme for linear regression, which guarantees the correctness of learning. In this article, to avoid malicious clouds from returning incorrect inference results, we design a privacy-preserving prediction scheme with lightweight verification. Our formal security analysis shows that GuardLR achieves privacy, completeness, and soundness. Empirical experiments using real-world datasets also demonstrate that GuardLR has high computational efficiency and accuracy.

Cervical spondylosis (CS) recognition systems provide regular screening services outside of a hospital and promote early detection and treatment of CS. However, in this paper, we propose a deep learning-based CS recognition system. Concerning the state-of-the-art and state-of-the-practice systems, the innovations of our approaches and algorithms are as follows: First, to elevate the reliance upon the sample number required for training the high-quality model, we reduce sample dimension and find optimal neural network architectures to reduce the number of model parameters to fit. Second, we incorporate multi-stream parallel network architecture search with multi-view feature extraction by converting time series classification into an image classification task. Specifically, five feature extraction methods (time-domain, frequency-domain, time-frequency domain, model-based, nonlinear feature extraction) are firstly utilized to extract features from multiple perspectives and form low-dimensional data set with multi-properties. Third, we reorganize low-dimensional data into image one representing the spatio-temporal relationship of muscle activity pattern. Finally, a multi-stream parallel network architecture search is proposed to use a bypass mechanism for optimal neural network architecture, each of which processes a kind of features mentioned above with an idea of the sparse connection of convolution neural network. The results on the real-world data set show that our CS recognition system achieves the average accuracy of 95.54%, average sensitivity of 99.09%, and average specificity of 90.00%, outperforming the state-of-the-art ones.

With the popularization of Internet of Things (IoT) devices in smart home and industry fields, a huge number of IoT devices are connected to the Internet. However, what devices are connected to a network may not be known by the Internet Service Provider (ISP), since many IoT devices are placed within small networks (e.g., home networks) and are hidden behind network address translation (NAT). Without pinpointing IoT devices in a network, it is unlikely for the ISP to appropriately configure security policies and effectively manage the network. Additionally, inferring fine-grained user interactions of IoT devices is also an interesting yet unresolved problem. In this paper, we design an efficient and scalable system via spatial-temporal traffic fingerprinting from an ISP's perspective in consideration of practical issues like learning-testing asymmetry. Our system can accurately identify typical IoT devices in a network, with the additional capability of identifying what devices are hidden behind NAT and the number of each type of device that share the same IP address. Our system can also detect user interactions and meanwhile identify their (concurrent) number through a multi-output regression model. Through extensive evaluation, we demonstrate that the system can generally identify IoT devices with an F1-Score above 0.999, and estimate the number of the same type of IoT device behind NAT with an average error below 5%. By studying 29 user interactions of 7 devices, we show that our system is promising in detecting user interactions.

We propose a dual-polarized leaky-wave antenna (LWA) with orthomode based on a metallic square waveguide, for broadband endfire radiation applications. Several double-sided slots acting as radiation elements are notched on the waveguide to produce an endfire pattern. To overcome the restriction associated with the endfire radiation condition of a linear array, a metallic core periodically loaded with pins is used to control the dispersion of the proposed double-sided-slot waveguide. By optimizing the height distribution of the loading pins along the longitudinal direction using a cosine function, stable endfire radiation can be achieved over a broad operating band. Moreover, dual-polarized radiation can be automatically implemented because of the rotational symmetry of the proposed LWA. Furthermore, two rectangular waveguides connected with a square waveguide are used as the orthomode transducer (OMT) of the proposed LWA. The numerical and experimental results verify that dual-polarized endfire radiation is achieved in the frequency range of 7.6-9.1 GHz with an isolation of higher than 45 dB. The realized gain achieved is in the range of 13-15 dBi with a cross-polarization level of below -30 dB.

A novel strategy to fabricate bilayer hydrogel actuators based on the asymmetric distribution of crystalline regions across the bilayer structures was proposed. By employing PVA polymer chains into an alkali solvent-derived chitosan hydrogel matrix, chitosan/PVA hybrid bilayer hydrogels with both excellent responsive bending and mechanical properties were obtained as pH-controlled manipulators. In the design, the chitosan/PVA hydrogels upon treatment with freeze-thawing cycles were taken as the first monolayer, where excessive crystalline regions appeared. The original chitosan/PVA hydrogel as the second monolayer was then integrated into one bilayer device through the chemical-crosslinking of epichlorohydrin at the interface. The results showed that the resultant chitosan/PVA bilayer hydrogel actuator with a weight ratio of 3 : 1 displayed better sensitivity upon exposure to stimuli. The actuation behaviors are strongly dependent on experimental parameters such as the pH, PVA content and the chemical-crosslinking density. It is proposed that the driving force originates from the asymmetric distribution of crystalline regions, thus resulting in differential swelling ratios between the monolayers. In addition, programmable 3D shape transformations were achieved by using the bilayer hydrogel with designed 2D geometric patterns, and the tailored gripper-like hydrogel actuator can successfully capture and transport the cargo. Moreover, this actuation behavior can be erased and re-written on demand under certain conditions. Taking advantage of this universal strategy, more attractive actuators derived from synthetic or natural polymers in combination with PVA are highly expected, which can be used as smart soft robots in various fields such as manipulators, grippers, and cantilever sensors.

Hyperspectral unmixing is recognized as an important tool to learn the constituent materials and corresponding distribution in a scene. The physical spectral mixture model is always important to tackle this problem because of its highly ill-posed nature. In this article, we introduce a linear spectral mixture model (LMM)-based end-to-end deep neural network named SNMF-Net for hyperspectral unmixing. SNMF-Net shares an alternating architecture and benefits from both model-based methods and learning-based methods. On the one hand, SNMF-Net is of high physical interpretability as it is built by unrolling Lp sparsity constrained nonnegative matrix factorization (Lp-NMF) model belonging to LMM families. On the other hand, all the parameters and submodules of SNMF-Net can be seamlessly linked with the alternating optimization algorithm of Lp-NMF and unmixing problem. This enables us to reasonably integrate the prior knowledge on unmixing, the optimization algorithm, and the sparse representation theory into the network for robust learning, so as to improve unmixing. Experimental results on the synthetic and real-world data show the advantages of the proposed SNMF-Net over many state-of-the-art methods.

This paper presents a novel framework of modeling crushable granular materials under mechanical loadings based on the discrete element method (DEM). The framework is featured with the construction of the one-to-one model in which every particle in a physical experiment has its own numerical twin and allows the modeling of irregular shaped fragments during the continuous breakage process. First, image processing techniques and spherical harmonic (SH) analysis were adopted, respectively, to segment and label particles and to construct a one-to-one model mathematically in DEM. Then, a particle crushing criterion based on the maximum interparticle contact force was used to predict the crushing events, showing fitting results that agreed very well with a large number of single particle crushing tests. Next, a statistical approach for the generation of particle fragmentation modes of a given type of sand particles based on the principal component analysis (PCA) was proposed. The aim of the PCA was to analyze the statistical trends of the coefficient matrix, which was composed of the SH coefficients of all the particles involved in the analysis. Finally, a successful modeling of a particle crushing event was achieved by replacing the particle, which was judged by the crushing criterion to undergo crushing, with a few subparticles chosen randomly from a specific fragment template constructed using the microcomputed tomography (micro-CT) data.

The immune checkpoint blockade therapy has profoundly revolutionized the field of cancer immunotherapy. However, despite great promise for a variety of cancers, the efficacy of immune checkpoint inhibitors is still low in colorectal cancer (CRC). This is mainly due to the immunosuppressive feature of the tumor microenvironment (TME). Emerging evidence reveals that certain chemotherapeutic drugs induce immunogenic cell death (ICD), demonstrating great potential for remodeling the immunosuppressive TME. In this study, the potential of ginsenoside Rg3 (Rg3) as an ICD inducer against CRC cells was confirmed using in vitro and in vivo experimental approaches. The ICD efficacy of Rg3 could be significantly enhanced by quercetin (QTN) that elicited reactive oxygen species (ROS). To ameliorate in vivo delivery barriers associated with chemotherapeutic drugs, a folate (FA)-targeted polyethylene glycol (PEG)-modified amphiphilic cyclodextrin nanoparticle (NP) was developed for co-encapsulation of Rg3 and QTN. The resultant nanoformulation (CD-PEG-FA.Rg3.QTN) significantly prolonged blood circulation and enhanced tumor targeting in an orthotopic CRC mouse model, resulting in the conversion of immunosuppressive TME. Furthermore, the CD-PEG-FA.Rg3.QTN achieved significantly longer survival of animals in combination with Anti-PD-L1. The study provides a promising strategy for the treatment of CRC., Graphical abstract A folate-targeted PEGylated cyclodextrin-based nanoparticle was developed for co-delivery of Rg3 and QTN. The combination of the resultant co-formulation and anti-PD-L1 antibody achieved chemo-immunotherapy for colorectal cancer.Image 1

In this work, we demonstrated that the co-crystallization strategy has offered an efficient and promising alternative route to achieve high-performance n-type semiconductors through charge-transport switching from pristine p-type systems. By using a simple “green synthesis” process through molecular “doping” with F4TCNQ into a p-type planar azaacene derivative TMIQ (0.27 cm2 V−1 s−1) host, charge transport characteristic switching occurs with a high electron mobility of 0.12 cm2 V−1 s−1 under atmospheric conditions obtained for the D–A complex TMF4TQ (cocrystal). The reasons for such switching lie in the ingenious energy level and molecular packing arrangement tailoring. Specifically, the insertion of F4TCNQ molecules has led to packing transformation from herringbone stacking (TMIQ) to a dense 2D brick arrangement and the low-lying LUMO levels (−4.55 eV) aligned to gold electrodes, thereby facilitating efficient electron injection and transport, and ensuring the air-stable nature, which is further confirmed using theoretical calculations. We believe that our work would provide new insights into high-performance air-stable n-type organic semiconductors exploration.

In the early morning on December 17, 2020 Beijing time, China's chang'E-5 probe successfully returned to the Earth with 1731 g of lunar samples after completing drilling, shoveling, packaging of lunar soil and scientific exploration on lunar surface. It is the successful completion of the third phase of China's lunar exploration project, namely “circling, landing and returning to the moon”. The scientific objectives of CE-5 mission are to carry out in situ investigation and analysis of the lunar landing region, laboratory research and analysis of lunar return samples. This paper analyzes scientific exploration tasks of CE-5 mission conducted on the lunar surface, and carries out the scientific payload system architecture design and individual scientific payload design with the scientific exploration task requirements as the target, and proposes the working mode and main technical index requirements of the scientific payloads. Based on the preliminary geological background study of the Mons Ruemker region which is the landing region of CE-5, the lunar scientific exploration and the laboratory physicochemical characterization of the return samples are of great scientific significance for our in-depth understanding of the formation and evolution of the Earth-Moon system and the chemical evolution history of the lunar surface.

In the era of cloud computing, to achieve convenient location-based service (LBS), consumers such as users, companies, and organizations prefer subcontracting massive geographical data to public clouds after encryption for privacy and security. However, numerous harmful cyber-attacks happen on those public clouds in an unpredicted and hourly manner. To alleviate those concerns, various secure query schemes on the encrypted data have been proposed in the literature. As a fundamental query of LBSs, forward-secure range query has not been well investigated. To address this issue, we propose a lightweight and forward-secure range query (LS-RQ) on geographically encrypted data, which soundly balances between security and efficiency. Promisingly, we design an index mechanism to manage geographical data on the public clouds, while not compromising the privacy of data. Moreover, our LS-RQ schemes provide a convenient approach to range query on geographically encrypted data on-the-fly. We also rigorously prove that LS-RQ is forward-secure. Finally, extensive experimental studies are performed on both real and synthetic datasets. By observation, our LS-RQ schemes are highly efficient in realistic environments. Particularly, on encrypted datasets with about 1, 000, 000 geographical data, our solution to secure range query takes strictly less than a second.

Nowadays, an increasing amount of data has been sent to the cloud for analysis and storage, and data security in the cloud has been widely concerned. Among them, CP-ABE is regarded as one of the most promising technologies to protect outsourced data. However, in most CP-ABE schemes, attackers may obtain user privacy information from policy plaintext. More likely, partial policy hiding programme neither satisfies full policy hiding nor applies to the needs of the large universe. In this paper,we hide both attribute values and names in policy by using private set intersection (PSI). The CP-ABE programme not only supports a complete hiding policy, but also can calculate the authorization relationship and mapping relationship between user passwords and keys. We use a polynomial-based PSI and a recursive algorithm to calculate the authorization relationship. And with the help of the algorithm and the label vector, the mapping is determined under communication restrictions. Through the outsourcing index, we achieve an efficient hiding strategy and effectively reduce the user's computing overhead. Finally, theoretical analysis and experiments show that our model has better performance while effectively protecting sensitive attributes.

Hyperspectral image (HSI) denoising is an ill-posed inverse problem. The underlying physical model is always important to tackle this problem, which is unfortunately ignored by most of the current deep learning (DL)-based methods, producing poor denoising performance. To address this issue, this paper introduces an end-to-end model aided nonlocal neural network (MAC-Net) which simultaneously takes the spectral low-rank model and spatial deep prior into account for HSI noise reduction. Specifically, motivated by the success of the spectral low-rank model in depicting the strong spectral correlations and the nonlocal similarity prior in capturing spatial long-range dependencies, we first build a spectral low-rank model and then integrate a nonlocal U-Net into the model. In this way, we obtain a hybrid model-based and DL-based HSI denoising method where the spatial local and nonlocal multi-scale and spectral low-rank structures are effectively exploited. After that, we cast the optimization and denoising procedure of the hybrid method as a forward process of a neural network and introduce a set of learnable modules to yield our MAC-Net. Compared with traditional model-based methods, our MAC-Net overcomes the difficulties of accurate modeling thanks to the strong learning and representation ability of DL. Unlike most “black-box” DL-based methods, the spectral low-rank model is beneficial to increase the generalization ability of the network and decrease the requirement of training samples. Experimental results on the natural and remote sensing HSIs show that MAC-Net achieves state-of-the-art performance over both model-based and DL-based methods. The source code and data of this article will be made publicly available at https://github.com/bearshng/mac-net for reproducible research.

Data deduplication technique has been widely adopted by commercial cloud storage providers, which is important in coping with the explosive growth of data. To further protect the security of users' sensitive data in the outsourced storage mode, many secure data deduplication schemes have been designed and applied in various scenarios. Among these schemes, secure and efficient re-encryption for encrypted data deduplication attracted the attention of many scholars, and many solutions have been designed to support dynamic ownership management. In this paper, we focus on the re-encryption deduplication storage system and show that the recently designed lightweight rekeying-aware encrypted deduplication scheme is vulnerable to the stub-reserved attack. Furthermore, we propose a secure data deduplication scheme with efficient re-encryption based on the convergent all-or-nothing-transform (CAONT) and randomly sampled bits from the Bloom filter. Due to the property of hash function, our scheme can resist the stub-reserved attack and guarantee the data privacy of data owners' sensitive data. Moreover, instead of re-encrypting the entire package, data owners are only required to re-encrypt a small part of it through the CAONT, thereby effectively reducing the computation overhead. Finally, security analysis and experimental results show that our scheme is secure and efficient in re-encryption.

Person re-identification (Re-ID) is a challenging research topic which aims to retrieve the pedestrian images of the same person that captured by non-overlapping cameras. Existing methods either assume the body parts of the same person are well-aligned, or use attention selection mechanisms to constrain the effective region of feature learning. But these methods concentrate only on coarse feature representation and cannot model complex real scenes effectively. We propose a novel Local Attention Guided Network (LAG-Net) to not only exploit the most salient area among different people, but also extract important local detail through a Local Attention System (LAS). LAS is an attention selection unit that could extract approximate semantic local features of human body parts without extra supervision. To learn discriminative attention feature representation, we explore an attention feature regularization scheme to enhance the relevance of body part features that belong to same personal identity. Considering the effectiveness of feature augmentation in the Re-ID task and the defect of the existing methods, we propose a Batch Attention DropBlock (BA-DropBlock) to further improve DropBlock by combining the attention selection mechanism. Results on mainstream datasets demonstrate the superiority of our model over the state-of-the-art. Especially, our approach exceeds the current best method by a large margin of 4.6 % on the most challenging dataset CUHK03.

With the rapid development of mobile networks and e-commerce, clothing recommendation has achieved considerable success in recent years. Fashion outfit matching has become an essential component to users while shopping, which helps users to select and present items to individuals in a personalized fashion recommendation. Apparently, it is an arduous task to guide complementary clothing matching due to the complexity and subjectivity of fashion items. Some existing solutions have been presented in recent years, which are tending to discover a series of visual cues to establish the matching relations. However, it would be mismatched easily due to these methods being hard to represent all the potential semantic information from the appearance of clothes. To thoroughly make use of the visual characteristics of clothing products and the related description information, we propose a complementary clothing matching method with some compatibility knowledge, named it CCMCK shortly. For visual compatibility, we adopt the graph neural network to model the visual relationship between items. To generate an outfit that satisfies the requirement of fashion compatibility, we propose a matching way under the compatibility constraint and seek to recommend compatible items based on multi-modal compatibility. Finally, we performed a qualitative investigation on the fill-in-the-blank and fashion outfit compatibility tasks to evaluate the proposed method.

Breast cancer is the leading cause of cancer death among women and almost all of the breast cancer-caused mortality is related to metastasis. It has been reported that glucocorticoid facilitates the metastasis of breast cancer in mice, and mifepristone can antagonize the effect of glucocorticoid. Paclitaxel is one of the important drugs in the treatment of breast cancer. Mifepristone combined with paclitaxel may be an effective strategy for anti-metastasis of breast cancer. However, the inherent defects of them in short blood circulation half-life and lack of tumor targeting not only limit their effectiveness but also cause adverse reactions. Therefore, our aim is to explore a novel protocol against breast cancer metastasis, further optimize its therapeutic efficacy by nano-delivery system, and initially explore its mechanism. Herein, a paclitaxel-conjugated and mifepristone-loaded hydrogel (PM-nano) was prepared by self-assembly. Its characterizations were studied. The anti-metastasis effect was also evaluated in vitro and in vivo. Its mechanism was also explored by western blot assay. The resulted PM-nano was developed with favorable water solubility and good biocompatibility. Moreover, PM-nano displayed increased cell uptake property and stimulated drug release in tumor micro-acidic environment. PM-nano was more effective in inhibiting the proliferation and metastasis of breast cancer than other groups in vitro and in vivo. PM-nano might inhibit metastasis through glucocorticoid receptor/receptor tyrosine kinase-like orphan receptor 1 and matrix metalloproteinases. Taken together, PM-nano exhibited superior anti-metastasis effect against breast cancer and excellent biocompatibility in vitro and in vivo, providing a new option for anti-metastasis.

With the advancement of mobile internet, cloud computing, and smart sensing devices, Location-Based Services (LBSs) have become more and more indispensable in the Internet of Things (IoT) based smart cities. Especially, spatial keyword queries have been widely deployed in real-life applications in recent years. Recently, several privacy-preserving spatial keyword queries schemes were proposed to guarantee data security and query privacy on outsourced data. However, these schemes support neither dynamic update nor diverse query types, which cannot meet the requirements in practical applications. This paper proposes two Secure Dynamic Spatial Keyword Queries (SDSKQ) constructions that support expressive query types and dynamic update. Firstly, we present a basic SDSKQ construction based on hidden-vector encryption and order-revealing encryption. Specifically, we propose a secure hybrid index structure for spatio-textual data, named Encrypted Textual Signature Quadtree (ETSQ-tree). Using ETSQ-tree, the server can prune the index tree according to search queries to reduce the search space. Besides, the ETSQ-tree can be updated dynamically. To resist the file-injection attack, which aims to infer query information according to newly inserted objects, we further improve the basic SDSKQ to achieve forward security. We implement our two constructions and evaluate them using real-world datasets. The experimental results show that they are efficient and feasible in practical applications, and the comparative evaluation confirms that the performance of our constructions outperforms that of the state-of-the-art schemes.

Systematic and dynamic humoral immune reconstitution is little-known for patients with relapsed/refractory (R/R) multiple myeloma (MM) who received anti–B-cell maturation antigen (BCMA) chimeric antigen receptor (CAR) T-cell therapy. We investigated the kinetics of B-cell, normal plasma cell, and immunoglobulin recovery in 40 patients who achieved ongoing response after anti-BCMA CAR T-cell therapy. All patients developed B-cell aplasia and the median duration of B-cell aplasia was 70 days (range, 23-270). The B-cell count reached its nadir on median day 7 and returned to baseline level on median day 97. BCMA+ cells in bone marrow turned undetectable on median day 28 (13-159) in 94.87% (37 of 39) of patients. Normal plasma cells in bone marrow were first redetected on median day 212. All patients developed a significant decrease in serum IgG, IgA, and IgM on median day 60. At year 1, recovery of serum IgG, IgM, and IgA was observed in 53.33% (8 of 15; non-IgG MM), 73.08% (19 of 26; non-IgM MM), and 23.81% (5 of 21;non-IgA MM) of the patients, respectively. Median time to IgG, IgM, and IgA recovery were days 386, 254, and not reached during follow-up, respectively. Virus-specific IgG levels decreased with loss of protection. Twenty-three of 40 (57.5%) patients had a total of 44 infection events. There were no infection-related deaths. These results reveal a 7-month aplasia of bone marrow normal plasma cells and longer period of hypogammaglobulinemia, suggesting a profound and lasting humoral immune deficiency after anti-BCMA CAR T-cell therapy, especially for IgA.

Vision-language navigation (VLN) is the task of navigating an embodied agent to carry out natural language instructions inside real 3D environments. In this paper, we study how to address three critical challenges for this task: the cross-modal grounding, the ill-posed feedback, and the generalization problems. First, we propose a novel Reinforced Cross-Modal Matching (RCM) approach that enforces cross-modal grounding both locally and globally via reinforcement learning (RL). Particularly, a matching critic is used to provide an intrinsic reward to encourage global matching between instructions and trajectories, and a reasoning navigator is employed to perform cross-modal grounding in the local visual scene. Evaluation on a VLN benchmark dataset shows that our RCM model significantly outperforms baseline methods by 10 percent on Success Rate weighted by Path Length (SPL) and achieves the state-of-the-art performance. To improve the generalizability of the learned policy, we further introduce a Self-Supervised Imitation Learning (SIL) method to explore and adapt to unseen environments by imitating its own past, good decisions. We demonstrate that SIL can approximate a better and more efficient policy, which tremendously minimizes the success rate performance gap between seen and unseen environments (from 30.7 to 11.7 percent).