Insulator defect detection in transmission lines has become increasingly important for the safe operation of power systems within Internet of Things (IoT)-enabled smart grid infrastructures. However, complex backgrounds, diverse defect characteristics, and limited computing resources at IoT edge nodes pose significant challenges to existing detection methods. In existing defect detection methods, the fixed receptive field of convolutional neural networks (CNNs) limits their ability to adapt to multi-scale defects. Defect detection methods based on the Transformer architecture offer advantages in global modeling, but they are limited by their inability to accurately locate defect locations and high computational complexity, hindering their deployment on resource-constrained unmanned aerial vehicle (UAV) platforms as intelligent IoT sensing devices. Therefore, this paper proposes an insulator defect monitoring framework based on dynamic adaptive perception and a local-global attention enhancement fusion network (DPL-DETR), effectively addressing the core challenges of real-time, intelligent insulator defect detection at the IoT edge. First, a dynamic perception scale-adaptive network (DSNet) is designed, which leverages a learnable dynamic convolution kernel mechanism to significantly enhance the model's adaptive perception capability for defects of varying scales while maintaining lightweight characteristics essential for edge deployment. Second, a precise attention spatial enhancement network (PASEN) is constructed to achieve accurate localization and feature enhancement of defect regions in complex background environments, while substantially reducing computational complexity to meet real-time IoT application requirements. In addition, a local-global attention fusion (LGAF) module is proposed, which employs a dual-branch collaborative mechanism to effectively fuse local detail features with global semantic information, thereby enhancing the model's capability to recognize defects of varying morphologies under diverse environmental conditions. Experiments on the Transmission Line-ID and OPDL datasets show that our proposed defect detection method DPL-DETR achieves an effective balance between performance and efficiency, with mAP50 reaching 98.25% and mAP50-95 reaching 68.74%. Furthermore, the number of parameters is reduced by nearly 50%, and the computational complexity decreases to 79.1 GFLOPs, facilitating efficient deployment on UAV-mounted IoT edge devices. The results of all evaluation metrics surpass the best existing detection methods, demonstrating the potential of DPL-DETR as a reliable intelligent perception solution for IoT-based power infrastructure monitoring.

Ensuring atomic execution of cross-shard transactions is a fundamental challenge for sharded blockchains, particularly in scenarios demand coordination across multiple shards. However, existing solutions either rely on on-chain coordination, leading to high communication overhead, or leverage secure hardware for off-chain execution, imposing strong trust assumptions and reducing general applicability. To this end, we propose RollShard, a sharded blockchain that integrates stateless off-chain mechanism to efficiently process multi-shard transactions (MSTs). In RollShard, each MST is abstracted into a transaction DAG by the Sequencer Shard to ensure the authenticity of the transaction content and the correctness of its execution order. Batched MSTs are dispatched to off-chain executors, each of which simulates transaction logic using a virtual zero-state model integrate with a hierarchical state-delta tree (HSDT). The HSDT employs a Merkle Sum tree to precisely capture batched MSTs’ impact on per-shard account states. Based on the HSDT, the executor generates the zero-knowledge proof to attest the correctness of each shard’s state changes and global value conservation. The resulting net state deltas are then optimistically committed to the relevant shards without cross-shard coordination, reducing intra-shard coordination. We design a game-theoretic incentive mechanism to ensure rational behavior of off-chain executors, showing that honest execution forms a Nash equilibrium under collateral staking. Experimental results based on a prototype deployed in a local area network demonstrate that ROLLSHARDsignificantly outperforms two baseline coordination models proposed in ByShard, namely the Linear and Distributed designs. Specifically, under high workload, RollShard improves throughput by 44.9% and 158%, and reduces cross-shard latency by 38.9% and 42.1%, compared to the Linear and Distributed models, respectively.

User decision-making behavior in recommender systems is jointly driven by a large number of underlying factors. Learning and revealing the representations of these latent factors can provide more robustness and interpretability. However, mining the latent intentions of user decision-making behaviors in existing multimodal recommendation studies faces the following two key challenges: (i) Modal Noise Pollution : In multimodal user intent modeling, inputs from individual modalities are inevitably corrupted by noise of varying severity. During message passing, a large proportion of irrelevant or even contradictory signals are propagated and injected into item representations, which impedes the model’s ability to achieve pure semantic alignment at the content level. (ii) User Intent Confounding : Real-world items naturally possess multiple attributes, with different attributes of the same item influencing distinct potential user intentions. However, in existing modeling designs for user intent, such intents are mapped onto user–item interaction labels of the same coarse granularity. This many-to-one mapping relationship between intentions and items leads to significant confounding and dilution of users’ fine-grained intentions. To address the above challenges, this work pays special attention to the implied user intent behind pure multimodal features. Specifically, we construct a dynamic, adaptive, multimodal intent disentanglement model. This model adopts a non-ID paradigm and mines the distribution of user intents in multimodal scenarios directly from users’ decision-making behaviors. A comprehensive experimental study on the Amazon dataset shows that the method is effective and provides a novel learning scheme for mining user intent in multimodal scenarios.

Blockchain-based Federated Learning (BCFL) has attracted considerable attention in the intelligent IoT domain for its privacy-preserving and decentralized characteristics. Depending on their applicable scenarios, BCFL frameworks are categorized into two types: synchronous and asynchronous. However, synchronous BCFL struggles with low efficiency in heterogeneous IoT environments, while asynchronous BCFL suffers from slow convergence speed. In additional, Both BCFL incur significant resource consumption from blockchain consensus mechanisms which is unrelated to federated learning tasks, leading to resource wastage and poor scalability, making them unsuitable for large-scale IoT networks. To address these challenges, we propose CoCFL, a novel BCFL framework utilizing multi-chain collaboration. CoCFL introduces two lightweight sub-chains: PoCFL-CChain and PC-CChain, based on different FL strategy. PoCFL-CChain uses a synchronous FL strategy for learning devices with similar performance to generate high-accuracy models, while PC-CChain adopts an asynchronous strategy for heterogeneous devices, which can improving training efficiency. CoCFL assigns devices to suitable sub-chains based on their performance to carry out FL tasks and aggregates the sub-chain models into a global model. This multi-chain collaboration strategy enhances model accuracy and convergence speed and significantly improves the scalability of BCFL. In additional, the consensus mechanisms in CoCFL sub-chains not only maintain the blockchain ledger but also handle FL-related tasks such as detecting poisoning attacks, assigning roles, and distributing incentives. This design not only improving the efficiency of BCFL, but also enhances learning security and ensuring fair incentives. Experiments show that CoCFL improves learning accuracy by 6% and efficiency by 18% over existing BCFL frameworks. It also demonstrates excellent scalability, with time consumption liner decreasing as sub-chains increase, and can withstand up to 40% of poisoning attacks while ensuring fair incentives.

The cold-start recommendation has been challenging due to the limited historical interactions for new users and new items. Recently, methods based on meta-learning and graph neural networks have been effective in this problem. However, these methods mainly focus on the missing user-item interactions in cold-start scenarios, overlooking the missing of user/item feature information, which significantly limits the quality and effectiveness of node embeddings. To address this issue, we propose a new method called Feature-Structure Adaptive Completion Graph Neural Network (FS-GNN), which is designed to tackle the cold-start problem by simultaneously addressing the missing feature and structure information in a bipartite graph composed of users and items. Specifically, we first design a trainable feature completion module that leverages the knowledge emergence abilities of large language models to enhance node embedding and mitigate the impact of missing features. Then, we incorporate a three-channel structure completion module to simultaneously complete the structures among users-users, items-items, as well as users-items. Finally, we adaptively integrate the feature and structure completion modules in an end-to-end fashion, so as to minimize cross-module interference when completing features and structures simultaneously. This generates more comprehensive and robust embeddings for users and items in recommendation tasks. Experimental results on multiple public benchmark datasets demonstrate significant improvements in our proposed FS-GNN in cold-start scenarios, outperforming or being competitive with state-of-the-art methods.

The widespread adoption of online recruitment platforms has highlighted the challenge of achieving efficient and personalized person-job fit. Existing approaches focus on interaction patterns or textual content but fail to leverage the rich relationships among candidates, thus limiting their ability to capture the full spectrum of candidates' characteristics. In this paper, we propose a novel model, FitCLM, which integrates graph representation learning and large language models (LLMs) to enable a comprehensive understanding of candidate features. FitCLM first constructs a candidate relationship graph based on historical recruitment data, uncovering potential connections and similarities among candidates. It then utilizes LLMs to process historical job-candidate matching data, producing fine seman-tic representations. Finally, through self-supervised dual-view ranking optimization, FitCLM facilitates collaborative learning between the interaction graph and the relationship graph, while cross-modal contrastive learning aligns graph-based and LLM-generated embeddings. Extensive experimental results demonstrate that FitCLM outperforms state-of-the-art collaborative filtering and content-based baseline methods on NDCG@5 and MRR@5 metrics across two real-world recruitment datasets, with notable advantages in data-sparse scenarios. This study offers new insights into achieving accurate and efficient person-job fit, highlighting the significant potential of LLMs for generating contextual information from recruitment data.

With the advancement of large-scale models, the demand for emotionally rich, long-context, and highly natural communication in human-computer interaction increases. However, the exploration of long-context or script-level speech conversation tasks remains limited due to the lack of specific supervised data. To address this, we introduce a three-stage data processing pipeline for creating a Chinese expressive long-dialogue speech dataset with scripts (CELSDS). We collect videos from TV series, manually annotate speaker information for each character, apply Optical Character Recognition (OCR) to extract speech content, annotate episode summaries, and use a large language model (LLM) to generate sentence-level scenario descriptions. To our knowledge, this is the first Chinese long-context dialogue dataset that incorporates speaker and content annotations, script-level episode summaries, and sentence-level scenario details. Using this dataset, we develop a baseline model for both speech-to-script and script-to-speech generation tasks. The annotations and data production code are open-sourced at: https://github.com/lijin0120/CELSDS.

In urban planning, the generation of urban environmental layouts is crucial for balancing functionality, aesthetics, and sustainability. However, existing studies mainly suffer from three major issues, namely "limited data resources," "neglect of multivariate features," and "lack of unified assessment indicators," which hinder the exploration of future data-driven urban environment layout planning methods. To this end, we propose LiPlan, the first large-scale multimodal dataset of livability layouts of urban environments. LiPlan covers 11,000 matched pairs of samples from 104 cities worldwide, captures multivariate urban features such as roads, buildings, green spaces, and water bodies, and provides a unified framework to assess livability from the perspectives of ecological land use and urban heat island effects. We validate the utility of LiPlan by applying two GAN-based generative models. Building on the unique characteristics of the dataset, we further propose a condition-driven two-stage generative model for more efficient and controllable high-quality layout generation. The dataset and related code are available at https://github.com/Seven-ed/LiPlan.

Smart contract vulnerability detection is an important task in securing the blockchain. However, existing detection methods primarily extract single view features, such as semantic or structural features, which ignores the synergistic supplementation of them to smart contract, remaining room for improvement in feature representation. To this end, this paper proposes the LLM-assisted dual-view awareness framework for smart contract vulnerability detection, which incorporates significantly different semantic features and structural features. To address the limitation of large language model (LLM) in domain-specific expertise, we design semantic awareness module based on Retrieval-Augmented Generation (RAG), construct vulnerability knowledge base, and perform semantic reasoning on smart contracts. To capture crucial structural information, we propose structural awareness module based on Graph Neural Network (GNN), construct contract graphs, and perform structural analysis on smart contracts. We evaluated four types of vulnerabilities, and the experimental results show that our approach significantly outperforms state-of-the-art approaches, achieving 4.80% improvement in accuracy for timestamp dependence detection.

Predicting pedestrian flow in scenic spots is a critical challenge for managing tourist areas. To address this, we propose the double-graph network (DGNet) framework, which combines occluded person re-identification and pedestrian flow prediction. Scenic spots are represented as nodes in a graph, with their pedestrian flow as node attributes, naturally forming a graph structure. DGNet consists of two key graphs: CNN-transformer graph (CTG) for occluded person re-identification and spatial-temporal graph (STG) for pedestrian flow prediction. CTG integrates global and local features using CNN, Transformer, and graph convolutional network (GCN) to handle occlusions effectively. STG employs spatial-temporal attention mechanisms to extract correlations across time and space for accurate pedestrian flow prediction. Based on comprehensive experiments, the proposed CTG obtains a comparable performance to the current mainstream occluded person re-identification algorithms. Comparative experiments with other models show that the STG achieves the best results on MAE, RMSE and MAPE metrics, outperforming other models by at least 0.29%, 0.51%, and 0.52%. These results highlight the framework’s robustness and accuracy. Moreover, the novelty of DGNet lies in its ability to bridge occluded person re-identification with flow prediction tasks, offering a scalable solution applicable to diverse scenic spots. This work provides practical insights into leveraging video surveillance for effective crowd management in tourist areas.

MLIR (Multi-Level Intermediate Representation) compiler infrastructure provides an efficient framework for introducing a new abstraction level for programming languages and domain-specific languages. It has attracted widespread attention in recent years and has been applied in various domains, such as deep learning compiler construction. Recently, several MLIR compiler fuzzing techniques, such as MLIRSmith and MLIRod, have been proposed. However, none of them can detect silent bugs, i.e., bugs that incorrectly optimize code silently. The difficulty in detecting silent bugs arises from two main aspects: (1) UB-Free Program Generation : Generates programs that are free from undefined behaviors to suit the non-UB assumptions required by compiler optimizations. (2) Lowering Support : Converts the given MLIR program into an executable form with a suitable lowering path that reduces redundant lowering passes and improves the efficiency of fuzzing. To address the above issues, we propose DESIL. DESIL enables silent bug detection by defining a set of UB-elimination rules based on the MLIR documentation and applying them to input programs. To convert dialects in the MLIR program into executable form, DESIL designs a lowering path optimization strategy to convert the dialects in the given MLIR program into executable form. Furthermore, DESIL incorporates the differential testing for silent bug detection. It introduces an operation-aware optimization recommendation strategy into the compilation process to generate diverse executable files. We applied DESIL to the latest revisions of the MLIR compiler infrastructure. It detected 23 silent bugs and 19 crash bugs, of which 17/16 have been confirmed or fixed.

In recent years, the Graph Transformer has demonstrated superiority on various graph-level tasks by facilitating global interactions among nodes. However, as for node-level tasks, the existing Graph Transformer cannot perform as well as expected. Actually, a node in a real-world graph does not necessarily have relationships with every other node, and this global interaction weakens node features. This raises a fundamental question: should we partition out an appropriate interaction channel based on graph structure so that noisy and irrelevant information will be filtered and every node can aggregate information in the optimal channel? We first perform a series of experiments on manually created graphs with varying homophily ratios. Surprisingly, we observe that different graph structures indeed require distinct optimal interaction channels. This leads us to ask whether we can develop a partitioning rule that ensures each node interacts with relevant and valuable targets. To overcome this challenge, we propose a novel Graph Transformer named Multi-channel Graphormer. The model is evaluated on six network datasets with different homophily ratios for the node classification task. Moreover, comprehensive experiments are conducted on two real datasets for the recommendation task. Experimental results show that the Multi-channel Graphormer surpasses state-of-the-art baselines, demonstrating superior performance.

Cooperative positioning (CP) is able to improve the vehicular positioning performance by introducing the data of multiple vehicles into the position estimation. However, CP methods are vulnerable to measurement outliers in dense urban areas. The existing outlier-resistant CP methods are easy to trap in local optimum and may wrongly reject the outliers when the ratio of outliers to inliers is relatively high. To deal with this problem, a factor graph optimization (FGO) based CP method using Graduated Non-Convexity (GNC) Welsch cost is proposed in this paper. The state-of-art FGO algorithm is used to integrate multi-node and multi-epoch measurements including the Global Navigation Satellite Systems (GNSS) pseudoranges, inter-epoch baselines estimated by GNSS time-differenced carrier phase (TDCP), and inter-vehicle ranging measurements in a centralized framework. The least-square cost in traditional FGO is replaced with the GNC-based Welsch cost so as to enhance the robustness of the proposed method to any kind of outliers in our CP system. The use of GNC can reduce the risk of local optimum by gradually increasing the non-convexity of the Welsch cost. The proposed method can de-weight the outliers correctly even if a large number of outliers exist. The experimental results show the superiority of the proposed method over the existing CP methods in resisting multiple outliers.

Self-supervised tasks show significant advantages for node representation learning in recommender systems. This core idea of self-supervised task-based recommender systems depends on data augmentation to generate multi-view representations. However, there are two key challenges that are not well explored in existing self-supervised tasks: i) Restricted by the structure of the graph-based CF paradigm itself, the classical graph comparison learning architecture ignores the global structural information on the user-item interaction graph. ii) In a key part of existing contrast learning-random graph data enhancement schemes can significantly deteriorate model performance. To address these challenges, we propose a new hypergraph collaborative filtering with adaptive augmentation framework(HCFAA). It captures both local and global collaborative relationships on the user-item graph through a hypergraph-enhanced joint learning architecture. In particular, the designed adaptive structure-guided model ignores the noise introduced on unimportant edges, and thus learns the critical node information on the user-item graph. Comprehensive experimental studies on the Amazon dataset show that the method is effective, which provides an optimization scheme with a new perspective for the problems of key node loss in graph data enhancement and loss of higher-order structural information in GNN. The source code of our model can be available onhttps://github.com/RSnewbie/RS/tree/master/HCFAA.

The application of knowledge graph completion in industry and the Internet of Things involves various aspects, ranging from improving production efficiency to providing intelligent decision support. During the pursuit of constructing a knowledge graph, information for the graph is obtained from textual documents and online web pages. The extraction of information from documents is incomplete due to the limited information, so the knowledge graph needs to be completed. The knowledge graph completion model maps triples into different vector spaces for representation, but it still has the following shortcomings: 1) the textual encoder lacks of structured knowledge, 2) the distribution of relationships is long-tailed, with the model predominantly predicting more frequent relationships. In this paper, we propose structure-augment based long-tailed knowledge graph completion model (SALT-KGC) to deal with the two issues. To tackle the first problem, we partition each triple into two asymmetric parts as in translation-based graph embedding approach. We encode both parts using a Siamese-style textual encoder. Our model employs both classifier and spatial measurement for representation and structure learning respectively, in order to increase the structured knowledge of encoders. To tackle the second problem, we implement the focal loss to solve the problem of imbalance between positive and negative samples and focus on hard examples. Moreover, we develop a self-adaptive ensemble scheme to further improve the performance by incorporating from an existing graph embedding model. The SALT-KGC model achieves state-of-the-art performance on two widely-used public datasets.

Phishing scams represent a significant criminal activity on Ethereum, driving the need for effective detection methods. The methods based on graph neural networks(GNNs) make significant breakthroughs due to their ability to model complex transaction networks. However, existing approaches often overlook the heterogeneity of Ethereum’s transaction graph during neighbor nodes aggregation. These methods typically focus on low-order neighbors, disregarding high-order ones, which limits their overall performance. To this end, we propose the High- and Low-order Transaction Aggregation Graph Network(HLTAG), which separately aggregates high- and low-order features for more effective feature representation. Specifically, we utilize biased random walk to aggregate low-order neighbors. We employ path aggregation to handle high-order neighbors. To mitigate the influence of noise and redundant information from high-order neighbors, we introduce a combination of attention decay, node similarity, and path attention mechanism, which dynamically adjust the aggregation weights. Extensive experiments demonstrate that HLTAG (94.4% Recall and 89.3% AUC) outperforms the state-of-the-art approaches in detecting Ethereum phishing scams, and exhibits significant advantages in large-scale scenarios.

One notable drawback of traditional Federated Learning (FL) is its susceptibility to single point of failures. In recent years, Blockchain-based Federated Learning (BCFL) has been proposed as an effective solution to address this issue. However, existing BCFL frameworks face challenges in heterogeneous IoT scenarios. The heterogeneity of IoT devices poses challenges to the adaptation of blockchain consensus. The integration of blockchain imposes constraints on the learning scalability of systems, making it challenging to accommodate a large number of heterogeneous IoT devices. On the other hand, current blockchain consensus fail to sufficiently measure the contributions and destructions among heterogeneous devices in terms of learning quality, leading to low learning security and insufficient incentive fairness. To overcome the limitations of prior art, this paper introduces CoCFL, a novel blockchain-based federated learning framework based on multi-chain collaborative model. CoCFL enhances learning scalability by adopting a multi-chain asynchronous collaboration approach that partitions both learning and communication granularity of the system. Within each sub chain, CoCFL introduces a lightweight, secure and incentive-fair blockchain-based federated learning consensus, called Proof of Contribution to FL (PoCFL). In PoCFL, partic-ipants' contributions to the learning and the consensus process form the basis for delegating consensus responsibility and dis-tributing rewards. Furthermore, we introduce a novel malicious model detection algorithm into PoCFL, called the Trustee Nearest Algorithm. Through Trustee Nearest, PoCFL effectively mitigates poisoning attacks. Experimental results demonstrate that CoCFL exhibits better learning scalability compared to traditional FL and and avdanced BCFL frameworks in the same scenarios and can effectively withstand poisoning attacks initiated by at least 40% of malicious participants. Moreover, CoCFL demonstrated good incentive fairness during the learning process.

An innovative architecture called execute-order-validate (EOV) has been proposed by Hyperledger Fabric that enables concurrent processing of transactions. However, the architecture suffers from issues such as excessive invalid transactions and serialization limitations in scenarios with high transaction conflicts, which restrict its applicability in real-time and high-performance settings. To address the aforementioned limitations, we propose ParFabric to enhance the EOV architecture. Firstly, we analyze four essential characteristics required for the transaction reordering algorithm within this architecture. We propose a heuristic dynamic reordering algorithm to reduce the number of invalid transactions. This is achieved through real-time identification and early abortion of transactions based on weighted pre-ordering and the construction of a transaction conflict graph. Secondly, leveraging the transaction conflict graph, we introduce a novel optimal block packing strategy based on transaction dependencies. This strategy replaces the total transaction order with partial order, enabling parallel validation and commit at the block level, thereby leading to increased system throughput while reducing transaction latency. Experimental results indicate that, ParFabric demonstrates excellent performance in terms of vertical scaling of peers. Additionally, at the same infrastructure cost, ParFabric provides 2.2x and 1.6x higher throughput than FabricPlusPlus and FabricSharp in high-conflict scenarios.

Blockchain systems widely employ Byzantine fault-tolerant (BFT) protocols to ensure consistency. Improving BFT protocols’ throughput is crucial for large-scale blockchain systems. Frontier protocols face crucial problems: (i) the binary dilemma between leader bottleneck in star-based linear communication and compromised resilience in tree-based sublinear communication; and (ii) 2- or 3-round protocols restrict the phase number of one proposal, thereby limiting the scalability and parallelism of the pipeline. To overcome the above problems, this paper proposes Crackle, the first sector-based pipelined BFT protocol with a sublinear communication complexity, for a throughput improvement of consensus protocol with max resilience of (N-1)/3. We propose a sector-based communication mode to disseminate messages from the leader to a subset of replicas in each phase to accelerate consensus and split the traditional two-round protocol into 2κ phases to increase the basic pipeline scale. When implementing Crackle, we address two technical challenges: (i) to ensure Quorum Certificate (QC) validation during continuous κ phases, we design a voteMap field within each block, and verify QC by the aggregation of continuous κ voteMaps; and (ii) to achieve pipeline decoupling among shorter phases, we propose a vote-appending mechanism that accelerates the leader’s transition to the next phase. We provide comprehensive theoretical proof of the correctness of Crackle, including safety and liveness. Moreover, we implement Crackle based on a public BFT framework and deploy it on 64 cloud servers. Real experimental results reveal that Crackle achieves up to 10.36x higher throughput compared with state-of-the-art BFT protocols such as Kauri and Hotstuff.

MLIR (Multi-Level Intermediate Representation) compiler infrastructure has gained widespread popularity in recent years. It introduces dialects to accommodate various levels of abstraction within the representation. Due to its fundamental role in compiler construction, it is critical to ensure its correctness. Recently, a grammar-based fuzzing technique (i.e., MLIRSmith) has been proposed for it and achieves notable effectiveness. However, MLIRSmith generates test programs in a random manner, which restricts the exploration of the input space, thereby limiting the overall fuzzing effectiveness. In this work, we propose a novel fuzzing technique, called MLIR. As complicated or uncommon data/control dependencies among various operations are often helpful to trigger MLIR bugs, it constructs the operation dependency graph for an MLIR program and defines the associated operation dependency coverage to guide the fuzzing process. To drive the fuzzing process towards increasing operation dependency coverage, MLIR then designs a set of dependency-targeted mutation rules. By applying MLIR to the latest revisions of the MLIR compiler infrastructure, it detected 63 previously unknown bugs, among which 38/48 bugs have been fixed/confirmed by developers.

Despite existing parallel chain systems improving blockchain throughput by allowing concurrent blocks to be appended, challenges such as the excessive number of waiting blocks before confirmation and the inconsistency between block generation order and global confirmation sequence still persist. To address these challenges, we propose CVchain, a parallel chain system with a cross-voting mechanism. Blocks from other subchains are incorporated into the consistency determination of the main chain, reducing the probability of confirmation errors. Our global sorting mechanism leverages both real-time height information and the implicit temporal order contained in voting to improve the accuracy of block ordering. Furthermore, our voting mechanism randomly splits the mining power of the system, preventing targeted attacks on the specific subchain and defending against liveness attacks. We prove the safety and liveness properties of CVchain. We demonstrated its performance with a prototype implementation and large-scale experiments involving 200 nodes across 10 cloud servers in a distributed network environment. The results indicate that CVchain achieves a latency reduction of approximately 32.3% at a confirmation error probability of 0.01 while maintaining throughput levels comparable to OHIE. Additionally, it provides enhanced transaction ordering services.

Genome-wide Association Studies (GWAS) methods have identified individual single-nucleotide polymorphisms (SNPs) significantly associated with specific phenotypes. Nonetheless, many complex diseases are polygenic and are controlled by multiple genetic variants that are usually non-linearly dependent. These genetic variants are marginally less effective and remain undetected in GWAS analysis. Kernel-based tests (KBT), which evaluate the joint effect of a group of genetic variants, are therefore critical for complex disease analysis. However, choosing different kernel functions in KBT can significantly influence the type I error control and power, and selecting the optimal kernel remains a statistically challenging task. A few existing methods suffer from inflated type 1 errors, limited scalability, inferior power or issues of ambiguous conclusions. Here, we present a new Bayesian framework, BayesKAT (https://github.com/wangjr03/BayesKAT), which overcomes these kernel specification issues by selecting the optimal composite kernel adaptively from the data while testing genetic associations simultaneously. Furthermore, BayesKAT implements a scalable computational strategy to boost its applicability, especially for high-dimensional cases where other methods become less effective. Based on a series of performance comparisons using both simulated and real large-scale genetics data, BayesKAT outperforms the available methods in detecting complex group-level associations and controlling type I errors simultaneously. Applied on a variety of groups of functionally related genetic variants based on biological pathways, co-expression gene modules and protein complexes, BayesKAT deciphers the complex genetic basis and provides mechanistic insights into human diseases.

Sharding is an important technology that utilizes group parallelism to enhance the scalability and performance of blockchain. However, the existing solutions use a historical transaction-based approach to reallocate shards, which cannot handle temporary overload and incurs additional overhead during the reallocation process. To this end, this paper proposes LMChain, an efficient load-migratable beacon-based sharding blockchain system. The primary goal of LMChain is to eliminate reliance on historical transactions and achieve the high performance. Specifically, we redesign the state maintenance data structure in Beacon Shard to effectively manage all account states at the shard level. Then, we innovatively propose a load-migratable transaction processing protocol built upon the new data structure. To mitigate read-write conflicts during the selection of migration transactions, we adopt a novel graph partitioning scheme. We also adopt a relay-based method to handle cross-shard transactions and resolve inter-shard state read-write conflicts. We implement the LMChain prototype and conducted experiments in a real network environment comprising 17 cloud servers. Experimental results show that, compared with state-of-the-art solutions, LMChain effectively reduces the average transaction wait latency of overloaded transactions by 30% to 48% in different cases within 16 transaction shards, while improving throughput by 3% to 10%.

The recognition accuracy of the pedestrian re-identification algorithm is affected by factors such as the scale, posture, occlusion level, and appearance of a pedestrian in an image. These factors bring great challenges to the further development of deep learning theory in the field of person re-identification. On the one hand, existing pedestrian re-identification model cannot accurately extract attention information containing contextual information and discriminative pedestrian features. On the other hand, images of the same pedestrian taken with different cameras have the problem of different degrees of occlusion. Therefore, we introduce an attention pooling mechanism, which allows the model to automatically focus on discriminative regions of pedestrian images. The model can then learn attention maps by automatically focusing on visually salient pedestrian regions. It can effectively address the influence of factors on the accuracy of pedestrian re-identification. Furthermore, we propose a pedestrian re-identification method based on saliency region detection and matching. The method first extracts multiple local saliency regions from pedestrian images using the pedestrian joint point detection method, then marks and extracts the positions of the abovementioned saliency regions, and finally performs matching and sorting. It can correct the impact caused by occlusion on the accuracy of pedestrian re-identification. We conduct tests and experiments on three public person re-identification datasets, and the results show that our method not only achieves the best recognition accuracy on the aforementioned datasets but also has better robustness.

Road traffic signs can improve the pressure of environmental traffic, and the real-time accurate recognition of traffic signs is conducive to the promotion and development of intelligent vehicles. However, in various complex application scenarios such as tilted and deformed traffic signs, there still have the following problems: First, the feature selection of the existing traffic recognition model only considers its network layer information and fails to retain the salient feature information with strong discriminatory power effectively, and the feature enhancement is easy to introduce background noise. Second, the existing traffic sign recognition models are difficult to deploy on mobile devices and systems, resulting in weak learning capability of feature representation in deep learning backbone network, and low robustness of the high-level feature fusion performance. Therefore, we propose a cascade attention mechanism, which can associate a series of attention units using a cascade approach, and design a cascade attention feature enhancement module, which can effectively improve the feature selection and feature enhancement performance in the traffic sign recognition process. Then, we design a lightweight deep learning model and propose modal fusion, a high-level feature-guided feature refinement mechanism, and a mutual attention enhancement module. In addition, the interrelationship between traffic signs and deep modal features is given, which can effectively enhance the feature representation learning ability of the lightweight deep learning model and significantly improve the efficiency and robustness of the model. The experimental results show that our proposed method achieves the most excellent recognition results on several traffic sign datasets.

The development of vehicle-to-vehicle (V2V) communication facilitates the study of cooperative positioning (CP) techniques for vehicular applications. The CP methods can improve the positioning availability and accuracy by inter-vehicle ranging and data exchange between vehicles. However, the inter-vehicle ranging can be easily interrupted due to many factors such as obstacles in-between two cars. Without inter-vehicle ranging, the other cooperative data such as vehicle positions will be wasted, leading to performance degradation of range-based CP methods. To fully utilize the cooperative data and mitigate the impact of inter-vehicle ranging loss, a novel cooperative positioning method aided by plane constraints is proposed in this paper. The positioning results received from cooperative vehicles are used to construct the road plane for each vehicle. The plane parameters are then introduced into CP scheme to impose constraints on positioning solutions. The state-of-art factor graph optimization (FGO) algorithm is employed to integrate the plane constraints with raw data of Global Navigation Satellite Systems (GNSS) as well as inter-vehicle ranging measurements. The proposed CP method has the ability to resist the interruptions of inter-vehicle ranging since the plane constraints are computed by just using position-related data. A vehicle can still benefit from the position data of cooperative vehicles even if the inter-vehicle ranging is unavailable. The experimental results indicate the superiority of the proposed CP method in positioning performance over the existing methods, especially when the inter-ranging interruptions occur.

Acoustic-to-articulatory inversion (AAI) aims to estimate the parameters of articulators from speech audio. There are two common challenges in AAI, which are the limited data and the unsatisfactory performance in speaker independent scenario. Most current works focus on extracting features directly from speech and ignoring the importance of phoneme information which may limit the performance of AAI. To this end, we propose a novel network called SPN that uses two different streams to carry out the AAI task. Firstly, to improve the performance of speaker-independent experiment, we propose a new phoneme stream network to estimate the articulatory parameters as the phoneme features. To the best of our knowledge, this is the first work that extracts the speaker-independent features from phonemes to improve the performance of AAI. Secondly, in order to better represent the speech information, we train a speech stream network to combine the local features and the global features. Compared with state-of-the-art (SOTA), the proposed method reduces 0.18mm on RMSE and increases 6.0% on Pearson correlation coefficient in the speaker-independent experiment. The code has been released at https://github.com/liujinyu123/AAINetwork-SPN.

Given one reference facial image and a piece of speech as input, talking head generation aims to synthesize a realistic-looking talking head video. However, generating a lip-synchronized video with natural head movements is challenging. The same speech clip can generate multiple possible lip and head movements, that is, there is no one-to-one mapping relationship between them. To overcome this problem, we propose a Speech Feature Extractor (SFE) based on memory-augmented self-supervised contrastive learning, which introduces the memory module to store multiple different speech mapping results. In addition, we introduce the Mixed Density Networks (MDN) into the landmark regression task to generate multiple predicted facial landmarks. Extensive qualitative and quantitative experiments show that the quality of our facial animation is significantly superior to that of the state-of-the-art (SOTA). The code has been released at https://github.com/Yaxinzhao97/MACL.git.

Given an audio clip and a reference face image, the goal of the talking head generation is to generate a high-fidelity talking head video. Although some audio-driven methods of generating talking head videos have made some achievements in the past, most of them only focused on lip and audio synchronization and lack the ability to reproduce the facial expressions of the target person. To this end, we propose a talking head generation model consisting of a Memory-Sharing Emotion Feature extractor (MSEF) and an Attention-Augmented Translator based on U-net (AATU). Firstly, MSEF can extract implicit emotional auxiliary features from audio to estimate more accurate emotional face landmarks. Secondly, AATU acts as a translator between the estimated landmarks and the photo-realistic video frames. Extensive qualitative and quantitative experiments have shown the superiority of the proposed method to the previous works. Codes will be made publicly available.

Owing to the attractive properties of decentralization, unforgeability, transparency, and traceability, blockchain is increasingly being used in various scenarios such as supply chain and public services, where massive Spatial-Temporal-Keywords (STK) transactions need to be packaged. However, due to the multi-dimensionality and randomness of STK transactions, existing solutions fail to enable queries in a verifiable and efficient way for blockchains storing multidimensional transactions. To this end, this article takes the first step to propose an authenticated and efficient query approach in hybrid blockchain systems consisting of on-chain and off-chain parts. We first design a data structure named MRK-Tree in the block body, which organizes STK transactions for efficient nodes pruning of both kNN and range queries. Then we propose an improved block header, which improves the efficient pruning of blocks on the basis of ensuring the authentication of query results. Also, we design a cross-block searching algorithm named Efficient Block Pruning (EBP) and intra-block searching algorithms named Authenticated kNN/Range Query (AKQ/ARQ) to accelerate authenticated queries for multiple MRK-Trees in the hybrid blockchain systems. Authentication mechanisms are proposed to ensure the soundness and completeness of query results. Rigorous security analysis validates the practicability of the proposed approach. We build a blockchain prototype to comprehensively evaluate the performance of proposed query schemes. Extensive evaluation results with real datasets reveal that our approach can ensure authenticated queries, meanwhile improving the time efficiency by up to 36.45x and space efficiency by up to 4 orders of magnitude compared with the well-known benchmark query schemes.

Acoustic-to-articulatory inversion (AAI) is to obtain the movement of articulators from speech signals. Until now, achieving a speaker-independent AAI remains a challenge given the limited data. Besides, most current works only use audio speech as input, causing an inevitable performance bottleneck. To solve these problems, firstly, we pre-train a speech decomposition network to decompose audio speech into speaker embedding and content embedding as the new personalized speech features to adapt to the speaker-independent case. Secondly, to further improve the AAI, we propose a novel auxiliary feature network to estimate the lip auxiliary features from the above personalized speech features. Experimental results on three public datasets show that, compared with the state-of-the-art only using the audio speech feature, the proposed method reduces the average RMSE by 0.25 and increases the average correlation coefficient by 2.0% in the speaker-dependent case. More importantly, the average RMSE decreases by 0.29 and the average correlation coefficient increases by 5.0% in the speaker-independent case.

Previous works derive personalized speech features by training the model on a large dataset composed of his/her audio sounds. It was reported that face information has a strong link with the speech sound. Thus in this work, we innovatively extract personalized speech features from human faces to synthesize personalized speech using neural vocoder. A Face-based Residual Personalized Speech Synthesis Model (FR-PSS) containing a speech encoder, a speech synthesizer and a face encoder is designed for PSS. In this model, by designing two speech priors, a residual-guided strategy is introduced to guide the face feature to approach the true speech feature in the training. Moreover, considering the error of feature’s absolute values and their directional bias, we formulate a novel tri-item loss function for face encoder. Experimental results show that the speech synthesized by our model is comparable to the personalized speech synthesized by training a large amount of audio data in previous works.

The rapid development of the blockchain leads to a blowout of on-chain transactions, contracts, and currencies, which will further accelerate the increase of data. The existing blockchain systems typically support exact transaction queries, which, however, cannot satisfy the QoS requirements with intelligent adjustment in the blockchain systems. To this end, this paper takes the first step to define and address the practically important problem of transaction cardinality estimation for QoS-adjustable intelligent blockchain systems. We first establish a mathematical relationship between the bit string and transaction cardinality. Thus, we can leverage the number of leading 1s of the obtained bit string to estimate the transaction cardinality. We then improve the block header and body with a corresponding search algorithm to access bit strings in blocks. We also propose an estimation protocol with intelligent adjustable QoS to support accuracy-guaranteed and efficiency-optimized estimation. Finally, we design an authentication scheme and guarantee the reliability of our protocol through rigorous theoretical derivation. When achieving the transaction cardinality estimation in blockchain, two technical challenges need to be addressed. (i) To ensure efficient, verifiable, and overhead-saving bit string accessing mechanism in blockchain, we propose the Merkle Cardinality Tree (MCT) and target block filtering mechanism based on Bloom Filter (BF) in off-chain and improve on-chain block header by joining the abstract of MCT and BF. (ii) To improve estimation efficiency while guaranteeing accuracy requirements in hybrid blockchain scheme, we propose a Dynamic One-round Sampling-based cardinality Estimation (DOSE) protocol and integrate BF-DOSE to intelligently accelerate estimation. We build MCT in Ethereum and store the MCT Root in the block header for estimation authentication. Extensive experiments reveal that our BF-DOSE protocol can well satisfy various accuracy and efficiency requirements of QoS-adjustable intelligent blockchain systems, and is one to two orders of magnitude faster compared with benchmark schemes.

With the continuous growth of the number of Internet of things(IoT) devices, more and more data are generated by IoT devices. IoT terminal devices need to transfer the data to the edge server for storage, so protecting the security of IoT data has become a huge challenge. As a new technology, blockchain has the characteristics of distributed, tamper-resistant. Smart contract running on the blockchain can automatically perform tasks. Because of these characteristics, blockchain can solve the data security problem of the IoT. In this paper, a blockchain-based data protection framework for the IoT in untrusted storage is proposed. Lightweight streaming authenticated data structures are also used to reduce the storage burden of the blockchain system and improve the efficiency of the framework. Finally, simulation results show that our work can reduce the storage burden of blockchain.

In recent years, crowdsourcing has become a new business model. Quality assessment of crowdsourcing has also become a hot topic of interest for researchers. Most quality control methods are based on centralised platforms and cannot guarantee complete reliability. Therefore, this paper proposes a blockchain-based quality assessment model for crowdsourcing(BC-CQAM). A trusted mechanism is introduced to construct a reputation model, based on that we also propose a blockchain-based worker selection algorithm(BC-WS). A new quality assessment algorithm(BIV-EM) is proposed, resulting in more accurate evaluation results. The validity of the algorithm and the reliability of the worker selection have been demonstrated experimentally.

In self-supervised monocular depth estimation, the depth discontinuity and motion objects' artifacts are still challenging problems. Existing self-supervised methods usually utilize two views to train the depth estimation network and use one single view to make predictions. Compared with static views, abundant dynamic properties between video frames are beneficial to refining depth estimation, especially for dynamic objects. In this work, we improve the self-supervised learning framework for depth estimation using consecutive frames from monocular and stereo videos. The main idea is to exploit an implicit depth cue extractor which leverages dynamic and static cues to generate useful depth proposals. These cues can predict distinguishable motion contours and geometric scene structures. Moreover, a new high-dimensional attention module is proposed to extract a clear global transformation, which effectively suppresses the uncertainty of local descriptors in high-dimensional space, resulting in a more reliable optimization in the learning framework. Experiments demonstrate that the proposed framework outperforms the state-of-the-art on KITTI and Make3D datasets.

Cued Speech (CS) is a communication system for deaf people or hearing impaired people, in which a speaker uses it to aid a lipreader in phonetic level by clarifying potentially ambiguous mouth movements with hand shape and positions.Feature extraction of multi-modal CS is a key step in CS recognition.Recent supervised deep learning based methods suffer from noisy CS data annotations especially for hand shape modality.In this work, we first propose a self-supervised contrastive learning method to learn the feature representation of image without using labels.Secondly, a small amount of manually annotated CS data are used to fine-tune the first module.Thirdly, we present a module, which combines Bi-LSTM and self-attention networks to further learn sequential features with temporal and contextual information.Besides, to enlarge the volume and the diversity of the current limited CS datasets, we build a new British English dataset containing 5 native CS speakers.Evaluation results on both French and British English datasets show that our model achieves over 90% accuracy in hand shape recognition.Significant improvements of 8.75% (for French) and 10.09% (for British English) are achieved in CS phoneme recognition correctness compared with the state-of-the-art.

Cued Speech (CS) is a visual communication system for the deaf or hearing impaired people. It combines lip movements with hand cues to obtain a complete phonetic repertoire. Current deep learning based methods on automatic CS recognition suffer from a common problem, which is the data scarcity. Until now, there are only two public single speaker datasets for French (238 sentences) and British English (97 sentences). In this work, we propose a cross-modal knowledge distillation method with teacher-student structure, which transfers audio speech information to CS to overcome the limited data problem. Firstly, we pretrain a teacher model for CS recognition with a large amount of open source audio speech data, and simultaneously pretrain the feature extractors for lips and hands using CS data. Then, we distill the knowledge from teacher model to the student model with frame-level and sequence-level distillation strategies. Importantly, for frame-level, we exploit multi-task learning to weigh losses automatically, to obtain the balance coefficient. Besides, we establish a five-speaker British English CS dataset for the first time. The proposed method is evaluated on French and British English CS datasets, showing superior CS recognition performance to the state-of-the-art (SOTA) by a large margin.

Genes do not function independently; rather, they interact with each other to fulfill their joint tasks. Identification of gene-gene interactions has been critically important in elucidating the molecular mechanisms responsible for the variation of a phenotype. Regression models are commonly used to model the interaction between two genes with a linear product term. The interaction effect of two genes can be linear or nonlinear, depending on the true nature of the data. When nonlinear interactions exist, the linear interaction model may not be able to detect such interactions; hence, it suffers from substantial power loss. While the true interaction mechanism (linear or nonlinear) is generally unknown in practice, it is critical to develop statistical methods that can be flexible to capture the underlying interaction mechanism without assuming a specific model assumption. In this study, we develop a mixed kernel function which combines both linear and Gaussian kernels with different weights to capture the linear or nonlinear interaction of two genes. Instead of optimizing the weight function, we propose a grid search strategy and use a Cauchy transformation of the P-values obtained under different weights to aggregate the P-values. We further extend the two-gene interaction model to a high-dimensional setup using a de-biased LASSO algorithm. Extensive simulation studies are conducted to verify the performance of the proposed method. Application to two case studies further demonstrates the utility of the model. Our method provides a flexible and computationally efficient tool for disentangling complex gene-gene interactions associated with complex traits.

Lip motion reflects behavior characteristics of speakers, and thus can be used as a new kind of biometrics in speaker recognition. In the literature, lots of works used two-dimensional (2D) lip images to recognize speaker in a text-dependent context. However, 2D lip easily suffers from various face orientations. To this end, in this work, we present a novel end-to-end 3D lip motion Network (3LMNet) by utilizing the sentence-level 3D lip motion (S3DLM) to recognize speakers in both the text-independent and text-dependent contexts. A new regional feedback module (RFM) is proposed to obtain attentions in different lip regions. Besides, prior knowledge of lip motion is investigated to complement RFM, where landmark-level and frame-level features are merged to form a better feature representation. Moreover, we present two methods, i.e., coordinate transformation and face posture correction to pre-process the LSD-AV dataset, which contains 68 speakers and 146 sentences per speaker. The evaluation results on this dataset demonstrate that our proposed 3LMNet is superior to the baseline models, i.e., LSTM, VGG-16 and ResNet-34, and outperforms the state-of-the-art using 2D lip image as well as the 3D face. The code of this work is released at https://github.com/wutong18/Three-Dimensional-Lip-Motion-Network-for-Text-Independent-Speaker-Recognition.

Smart vehicles can cooperate in teams to perform crowdsensing tasks in smart cities. A critical challenge in this regard is to build a secure model for nondeterministic vehicle teams to achieve maximum social welfare. Although several crowdsensing models have been proposed, none of them has focused on real-time vehicle teamwork. In this article, to the best of our knowledge, we propose the first secure model, called blockchain-based nondeterministic teamwork cooperation (BNTC), for nondeterministic teamwork cooperation in a vehicular crowdsensing system. We model the system as a multiconditional NP-complete problem by explicitly considering the dynamic features of task issuers and workers. To solve the problem, we propose the winning teams selected (WTS) algorithm based on a reverse auction and utilize a knapsack-based method to solve the models. We consider the credit of teams for determining the payment. Thus, we propose a credit-based team payment (CTP) algorithm for BNTC to maximize the welfare of the system. We also propose a general blockchain-based framework to address trust issues and security challenges to make the method suitable for use in practical applications. Based on theoretical analyses and extensive simulations, we demonstrate that the proposed model performs better than the baselines and can achieve the maximum social welfare. Implementation with Ethereum suggests our model can operate within a reasonable cost.

Image segmentation based on deep learning has greatly promoted the development of the field of computer-aided diagnosis. However, the large scale medical annotation of ground truth is so difficult that it directly affects the performance of existing segmentation models. In this work, an Attention based Semi-supervised Neural Network is proposed, which can complete end-to-end segmentation task of thyroid ultrasound image with weakly annotated classification data and a small amount of fully annotated segmentation data. Two kinds of attention modules are proposed to improve network performance through the trainable feedforward structure of bottom-up and top-down so as to suppress or activate the feature channels and image regions respectively. The experimental results show that when there is only 13% of fully annotated data, the Jaccard similarity coefficient of thyroid nodule segmentation is 74.91%, 4.97% higher than VGG-based semi-supervised model. The classification accuracy of benign and malignant is increased from 91.67% to 95.00%. Equally important, with the same number of fully annotated data, our model has better generalization than that of the supervised segmentation models.

For delay-sensitive applications in mobile edge computing (MEC), task admission approach is of vital importance, and there has been a lot of researches in this field. But previous works focus on the situation with only one MEC server that is a simplification of the real world. In multi-servers situation, some devices may be within the service range of multiple MEC servers, so that they could choose which MEC server to offload. We formulate this problem to a multiple-choice integer program (MCIP) and utilize Ben's genetic algorithm to solve it. The simulation results show that our approach can significantly reduce energy consumption and every task can catch its deadline under almost all experiments.

The incumbent sharding schemes usually assign the nodes to different committees randomly to meet the demands of security and efficiency at the same time. For example, Elastico protocol obtains a random value by letting the node perform proof of work, and then uses this value for sharding. However, the strategy of random sharding ignores the objective differences between nodes, causing performance gaps between different committees in blockchain. This creates a bottleneck in the transaction throughput of the blockchain. In the paper, we propose a node rating based sharding scheme for blockchain system called NRSS. The key idea of NRSS is to evaluate nodes in the network by both the speeds and results of transactions verification before, and then assign them into different committees by balancing the score to reduce the performance gap between committees and increase the speed of transaction process. We implement NRSS in a local blockchain system, and the experiment results show that NRSS can increase the sharding effect of a blockchain, with an average throughput increase of 32.2% in the simulation environment where the node performance difference is up to 75%, depending on the number of nodes in the committee that are preset in the blockchain.

In this paper, we investigate a nonuniform node distribution strategy to mitigate the energy hole problem in wireless sensor networks (WSNs). Firstly, based on the analysis of energy consumption, we deduce a novel continuous node density function. Secondly, with the transformation of the density function, we obtain the disk radius function. And then, based on the disk radius function, we propose a novel nonuniform node distribution using adaptive Poisson disk (NDAPD). Our strategy can make the node density vary continuously in the network and mitigate the energy hole problem effectively. Finally, a routing algorithm tailored is presented for the proposed nonuniform node distribution. Compared with other well-known nonuniform node distribution strategies, NDAPD can reduce the number of network nodes effectively, improve the utilization of network energy by 3%-5% and increase the data transmission rate to 100% in most cases. Like other strategies, our strategy can be applied to most scenarios.

BACKGROUND: Infections by RNA viruses such as Influenza, HIV still pose a serious threat to human health despite extensive research on viral diseases. One challenge for producing effective prevention and treatment strategies is high intra-species genetic diversity. As different strains may have different biological properties, characterizing the genetic diversity is thus important to vaccine and drug design. Next-generation sequencing technology enables comprehensive characterization of both known and novel strains and has been widely adopted for sequencing viral populations. However, genome-scale reconstruction of haplotypes is still a challenging problem. In particular, haplotype assembly programs often produce contigs rather than full genomes. As a mutation in one gene can mask the phenotypic effects of a mutation at another locus, clustering these contigs into genome-scale haplotypes is still needed. RESULTS: We developed a contig binning tool, VirBin, which clusters contigs into different groups so that each group represents a haplotype. Commonly used features based on sequence composition and contig coverage cannot effectively distinguish viral haplotypes because of their high sequence similarity and heterogeneous sequencing coverage for RNA viruses. VirBin applied prototype-based clustering to cluster regions that are more likely to contain mutations specific to a haplotype. The tool was tested on multiple simulated sequencing data with different haplotype abundance distributions and contig sizes, and also on mock quasispecies sequencing data. The benchmark results with other contig binning tools demonstrated the superior sensitivity and precision of VirBin in contig binning for viral haplotype reconstruction. CONCLUSIONS: In this work, we presented VirBin, a new contig binning tool for distinguishing contigs from different viral haplotypes with high sequence similarity. It competes favorably with other tools on viral contig binning. The source codes are available at: https://github.com/chjiao/VirBin .

This study describes a nonlinear geometric tongue model based on MRI and Cone-beam CT (CBCT) data. Comparing with the conventional geometric tongue model, the proposed tongue model is controlled by several prototype vertices, and the relationship between tongue mesh vertices and prototype vertices are modeled with quadratic functions. The results indicate that: i) quadratic models do improve the reconstruction performance of tongue mesh, especially in the tongue root region; ii) the quadratic model which use the cross-prototype-vertex information achieves the best performance of tongue mesh reconstruction; iii) the reconstruction performance can be further improved if an extra prototype vertex TP in the tongue root region is taken into account, even if TP is estimated from the measured prototype vertices.

Lip-reading has been successfully demonstrated that it can improve the performance of automatic speech recognition system especially in the presence of acoustic noise. However, the information about lip movement is still insufficient as the lip features are obtained from discrete three-dimensional points and planar images. The internal mechanisms of lip movement are not described and reflected. In this paper, we employed a novel deepening technique, namely densely connected convolutional networks (DenseNets), to obtain visual representation from color images. In addition, a new 3D lip physiologic feature based on the position and structure of facial muscles was extracted to represent the similarity of the way people speak. The color image feature and 3D lip geometric-physiologic feature were coupled together in the last fully-connected layer of DenseNets. The experimental results show that DenseNets can handle spatial-temporal information of a whole image sequence and the lip feature integrating our proposed 3D geometric-physiological feature is sufficient to improve the recognition rate by as much as 3.91% (from 94.84%, with the color images only, to 98.75%).

The movement of tongue plays an important role in pronunciation. Visualizing the movement of tongue can improve speech intelligibility and also helps learning a second language. However, hardly any research has been investigated for this topic. In this paper, a framework to synthesize continuous ultrasound tongue movement video from speech is presented. Two different mapping methods are introduced as the most important parts of the framework. The objective evaluation and subjective opinions show that the Gaussian Mixture Model (GMM) based method has a better result for synthesizing static image and Vector Quantization (VQ) based method produces more stable continuous video. Meanwhile, the participants of evaluation state that the results of both methods are visual-understandable.

MOTIVATION: It has been suggested that presumably distinct classes of genomic regulatory elements may actually share common sets of features and mechanisms. However, there has been no genome-wide assessment of the prevalence of this phenomenon. RESULTS: To evaluate this possibility, we performed a bioinformatic screen for the existence of compound regulatory elements in the human genome. We identified numerous such colocated boundary and enhancer elements from human CD4(+) T cells. We report evidence that such compound regulatory elements possess unique chromatin features and facilitate cell type-specific functions related to inflammation and immune response in CD4(+) T cells.

SUMMARY: Although some histone modification chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq) signals show abrupt peaks across narrow and specific genomic locations, others have diffuse distributions along chromosomes, and their large contiguous enrichment landscapes are better modeled as broad peaks. Here, we present BroadPeak, an algorithm for the identification of such broad peaks from diffuse ChIP-seq datasets. We show that BroadPeak is a linear time algorithm that requires only two parameters, and we validate its performance on real and simulated histone modification ChIP-seq datasets. BroadPeak calls peaks that are highly coincident with both the underlying ChIP-seq tag count distributions and relevant biological features, such as the gene bodies of actively transcribed genes, and it shows superior overall recall and precision of known broad peaks from simulated datasets. AVAILABILITY: The source code and documentations are available at http://jordan.biology.gatech.edu/page/software/broadpeak/.

MOTIVATION: Chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq) is widely used in biological research. ChIP-seq experiments yield many ambiguous tags that can be mapped with equal probability to multiple genomic sites. Such ambiguous tags are typically eliminated from consideration resulting in a potential loss of important biological information. RESULTS: We have developed a Gibbs sampling-based algorithm for the genomic mapping of ambiguous sequence tags. Our algorithm relies on the local genomic tag context to guide the mapping of ambiguous tags. The Gibbs sampling procedure we use simultaneously maps ambiguous tags and updates the probabilities used to infer correct tag map positions. We show that our algorithm is able to correctly map more ambiguous tags than existing mapping methods. Our approach is also able to uncover mapped genomic sites from highly repetitive sequences that can not be detected based on unique tags alone, including transposable elements, segmental duplications and peri-centromeric regions. This mapping approach should prove to be useful for increasing biological knowledge on the too often neglected repetitive genomic regions. AVAILABILITY: http://esbg.gatech.edu/jordan/software/map CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

A networked integrated manufacturing system oriented product lifecycle based on multi-agent is presented, in order to meet the informatics of enterprises, solve the problem of ldquoinformation isletrdquo in manufacturing, improve design level and manufacturing quality, shorten development and production period and reduce production and service costs. The functional modules and system architecture are studied, key technologies including architecture of the networked integrated manufacturing agents, system integration, data sharing and transformation, task assignment and request information ranking of technical services are researched and a three-layer network configuration is structured based on multi-agent. Based on these, a prototype system of networked integrated manufacturing system oriented product lifecycle based on multi-agent is established. Running result of the system proves that the theories and technologies are effective.

The customer satisfaction index is applied to collaborative negotiation in collaborative design environment. The index system of customer satisfaction in the process of collaborative design is analyzed. The negotiation model and negotiation process are studied. Based on these, an intelligent negotiation model is built by BP neural network. Combining with the actual demands of enterprise, training and simulating of the model are studied. Results show that the BP neural network model is effective. After training, the model can give impersonal evaluation to design team which takes part in product design. And the evaluation result has the directive significance to improve design and enhance customer satisfaction index.

On the analysis of current state of aeroengine remote diagnosis, collaborative mechanism based on multi-agent was introduced to overcome the obstacles of conventional remote fault diagnosis. The model of aeroengine remote collaborative diagnosis based on multi-agent was put forward on analysis of the positional relationship of all agents in the collaborative environment and the relationship between collaborative agents and roles in the course of collaboration. Some key technologies such as coordination mechanism, task assignment mechanism, agent interaction mechanism, case-based reasoning (CBR) in treatment agent, and the analytic hierarchy process (AHP) in decision analysis were discussed and specific methods of realization were given concretely. Based on these, a Web-based prototype system for aeroengine fault diagnosis was developed on the JADE (Java Agent DEvelopment Framework) platform. The process of system implementation and a case example of fault diagnosis were presented to illustrate and prove the proposed system's applicability. Running results show the feasibility and reliability of the framework, which will be helpful to integrate the aeroengine diagnosis knowledge, improve the diagnosis efficiency effectively and decrease the aeroengine diagnosis cost remarkably.