The multi-path commodity flow problem (MPCFP) is crucial for ensuring reliable and high-speed data transmission in communication networks. However, existing studies that employ pre-generated routing paths neglect real-time load state and the coupling among decisions, thus hindering the achievement of high-quality solutions. To overcome this, we propose Hierarchical Reinforcement Learning with Topology-Aware Exploration (HRL-TAE), which is the first fully end-to-end framework that dynamically produces high-quality solutions based on real-time network states. HRL-TAE integrates an exploration mechanism and utilizes the State Transition Guiding List (STGL) to guide state transitions, thereby transforming topology exploration into a Markov decision process. Guided by STGL, two closely coupled layers in HRL-TAE, that is, the path construct layer and the ratio allocate layer, construct multiple subpaths for each flow and allocate traffic ratios among them. Subsequently, adaptive constraint-driven masks exclude infeasible actions during decision making, thereby guaranteeing that all constraints are satisfied. We also adopt a tailored training approach to obtain accurate gradient estimates and improve training efficiency. Simulations and real-world experiments demonstrate that HRL-TAE achieves superior performance.

Abstract With the continuous integration of functions in wearable devices, power consumption demands have increased significantly, posing serious challenges to conventional power supply methods. Wearable self-powered technologies offer an effective solution to this issue. This study focuses on the efficient utilization of solar energy from the human wrist and presents the design and implementation of a flexible photovoltaic wristband. The wristband employs a multi-directional parallel array of photovoltaic cells, integrated with an energy management module, enabling it to adapt effectively to the dynamic and non-uniform solar irradiance conditions on the wrist. Through both simulated sunlight and real outdoor environment tests, the energy harvesting and load-driving performances of the photovoltaic wristband were comprehensively evaluated. The results show that under a highest average outdoor illuminance of 37.38 × 10 3 lx (525 W·m −2 ), the wristband delivers an average output power of 15.88 mW, providing a stable 3.3 V supply to a wearable motion sensing node, thereby enabling self-powered operation. During a complete “energy accumulation-load activation” cycle, the sensing node can operate for 37.84 s to perceive and transmit data. By employing a one-dimensional convolutional neural networks (1D-CNN) algorithm, accurate recognition of four motion states is successfully achieved. This work presents a systematic study covering energy harvesting scenarios analysis, wristband design, performance evaluation, and sensing application. The proposed photovoltaic wristband demonstrates excellent cyclic energy accumulation and stable power supply capabilities, validating the feasibility and practicality of the wearable photovoltaic self-powered system and highlighting its promising potential for future wearable applications.

Online anomaly detection (OAD) plays a pivotal role in real-time analytics and decision-making for evolving data streams. However, existing methods often rely on costly retraining and rigid decision boundaries, limiting their ability to adapt both effectively and efficiently to concept drift in dynamic environments. To address these challenges, we propose DyMETER, a dynamic concept adaptation framework for OAD that unifies on-the-fly parameter shifting and dynamic thresholding within a single online paradigm. DyMETER first learns a static detector on historical data to capture recurring central concepts, and then transitions to a dynamic mode to adapt to new concepts as drift occurs. Specifically, DyMETER employs a novel dynamic concept adaptation mechanism that leverages a hypernetwork to generate instance-aware parameter shifts for the static detector, thereby enabling efficient and effective adaptation without retraining or fine-tuning. To achieve robust and interpretable adaptation, DyMETER introduces a lightweight evolution controller to estimate instance-level concept uncertainty for adaptive updates. Further, DyMETER employs a dynamic threshold optimization module to adaptively recalibrate the decision boundary by maintaining a candidate window of uncertain samples, which ensures continuous alignment with evolving concepts. Extensive experiments demonstrate that DyMETER significantly outperforms existing OAD approaches across a wide spectrum of application scenarios.

Accurate process and measurement noise covariances are indispensable for Kalman filtering, yet they are difficult to identify when the innovation used for calibration is intermittently corrupted by sensor outliers. Conventional autocovariance least-squares (ALS) estimates the noise covariances by matching empirical and theoretical innovation autocovariances, but its quadratic criterion can convert a few impulsive autocovariance entries into large covariance bias. This paper develops an outlier robust ALS estimator, ALS-IRLS, for covariance identification under such contamination. ALS-IRLS casts ALS as a structured Huber M-estimation problem on the stacked multi-lag autoco variance equations and solves the induced weighted least-squares subproblems by iteratively reweighted least squares (IRLS) within the Riccati/gain fixed-point loop. This design preserves the information that separates process and measurement noise while bounding the influence of contaminated equations. Bounded influence, threshold-sensitivity, local-convergence, admissibility, and amortized-complexity analyses are provided. Simulations under 15% contamination show that ALS-IRLS reduces covariance estimation errors by over two orders of magnitude relative to ALS and yields downstream filtering accuracy close to the oracle Kalman filter, including against adaptive robust-filter baselines.

Operating large multi-reservoir systems is critical for effective water allocation, hydropower generation, and economic development. However, conventional robust planning-based methods scale poorly beyond single-reservoir or cascaded systems due to computational intractability. Addressing inter-reservoir relation extraction and coordination under conflicting objectives, uncertain inflows, and complex network couplings therefore remains an open challenge. To tackle this problem, we introduce the multi-agent reservoir cooperative operation (MARCO) environment, which integrates multiple objectives, heterogeneous topology, and stochastic inflows in a unified framework. An algorithm is then designed to construct a multi-relational directed acyclic graph (MR-DAG) that encodes the underlying topology through coupled objectives and entity relations. Building on this representation, we propose the multi-agent relational directed acyclic graph transformer (MAR-DAGT), a reinforcement learning algorithm that performs typed message passing for efficient feature extraction and employs acyclic decision-making to exploit causal structure for improved credit assignment. Extensive experiments on MARCO show that MAR-DAGT consistently outperforms other MARL and optimization-based baselines in terms of objective satisfaction and robustness to inflow uncertainty.

The inverse dynamic games problem is to model expert demonstrations by identifying the underlying cost functions of multiple agents from observed trajectories of their dynamic game interactions. This article investigates discrete-time, finite-horizon linear-quadratic (LQ) problems where both the state weight matrix and input weight matrix are unknown, with the presence of both process noise and observation noise. In addition, each player's cost function incorporates a player-specific, unknown linear term with respect to the state. Under this framework, first, sufficient conditions are established for the solvability of the weight matrices. Subsequently, it is proved that the inverse dynamic games problem involving heterogeneous unknown target states is structurally identifiable, unaffected by process noise. Building on the necessary conditions for Nash equilibrium solutions in forward problems, the estimation of the cost function parameters is formulated as a nontrivial solution to a homogeneous linear estimation problem, which can be implemented in a distributed manner. Furthermore, the proposed estimator achieves statistical consistency under the influence of observation noise. The effectiveness is illustrated through a multivehicle spring-coupled dynamic game and an interactive steering control scenario.

In this article, we investigate the distributed optimization problem of heterogeneous general linear multiagent systems by the adaptive dynamic programming (ADP) approach over directed communication networks. A distinctive feature of this work is the development of a data-driven approach that eliminates the need for prior knowledge of system dynamics for all agents. To address the challenges posed by unknown system dynamics, we utilize the ADP-based data-driven approach to develop the distributed optimization control law. First, the feedback gain of the control law is determined based on the state and input data of the controlled systems. Next, the system dynamics are reconstructed using the solved feedback gain and the running data of the controlled systems. Then, the remaining parameters in the control law are designed by solving a series of steady-state equations. Under standard assumptions and through the application of the certainty equivalence principle, we prove that the proposed approach solves the distributed optimization problem, ensuring output consensus of all agents at the optimal solution of the global cost function. Finally, the viability of our proposed approach is demonstrated through its application to optimal output power sharing control of hydraulic turbine systems and their large-scale form.

This article develops a scheme to tackle the safe optimal formation tracking issue for multiple fixed-wing uncrewed aerial vehicles (UAVs) with external disturbances and asymmetric control constraints. To ensure safety constraints in collision avoidance, a safe set is first constructed by a super level set of a continuously differential function, following a novel control barrier function (CBF) to characterize the safety. Subsequently, we transform the safe optimal formation tracking control into a constrained zero-sum (ZS) differential game to mitigate the destabilizing effects of the disturbances, where the cost function is constructed in a nonquadratic form to cope with asymmetric input constraints. Particularly, the designed CBF is integrated into the cost function to penalize the unsafe behavior, and a damping coefficient is included to balance the optimality and safety. Afterwords, a critic-only reinforcement learning (RL) strategy is developed to learn the robust safe Nash policy, where the critic weights are updated by applying experience replay technology, thus avoiding the requirement for persistence of excitation condition. Moreover, the stability and forward invariance of the safe set of the presented scheme are also verified. Finally, simulation examples are provided to substantiate the validity of the control scheme.

Multi-robot task planning (MRTP) at scale in robotic mobile fulfillment systems (RMFS) remains a challenge due to the curse of dimensionality and complex dynamic properties. Aiming to solve these challenges, we construct an end-to-end scalable multi-robot task planner capable of scaling to large-scale systems by learning hierarchical planning policies. In this planner, we design a centralized hierarchical temporal task planning framework to mitigate the curse of dimensionality while ensuring timely dynamic response. Following this framework, we propose a novel cycle-constrained asynchronous temporal graph (CycATG) to provide foundation for modeling the system dynamics. Based on the graph representation, we formulate the MRTP problem as a semi-Markov decision process (SMDP) that focuses solely on critical interaction points to improve computational and sampling efficiency. The policies in SMDP are parameterized via a hierarchical temporal attention network with temporal embedding layers to enhance spatio-temporal feature extraction. Additionally, the decoder masks in this network naturally ensure that the generated actions strictly satisfy the required dynamic hard constraints. The above hierarchical policies are jointly optimized using an efficient hierarchical REINFORCE with rollout counterfactual baseline method. To further enhance generalization performance on unlearned instances while preventing catastrophic forgetting, we extend it with region expansion curricula. Experiments demonstrate that our planner outperforms state-of-the-art methods on different MRTP instances across simulated and real-world RMFS. It successfully scales to instances with up to 200 robots, 1000 retrieval racks on unlearned maps while maintaining performance advantages.

Merge trees are fundamental topological descriptors with broad theoretical and practical significance, yet their computation, particularly for augmented merge trees, incurs substantial computational and memory overhead. We present BiGMT, a highly scalable parallel algorithm for merge tree construction based on regions of the Morse-Smale segmentation. By exploiting boundary vertex characteristics and the dynamic merging of these regions, BiGMT preserves the essential topology of d-dimensional manifolds and represents the structure as a weighted bipartite graph. This formulation decomposes merge tree construction into fine-grained parallel tasks, significantly reducing traversal cost and memory consumption. Furthermore, BiGMT efficiently constructs augmented merge trees using reverse binary lifting search and parallel connection mechanisms, substantially improving augmentation performance. Extensive experiments on medium- to large-scale scalar field datasets demonstrate substantial GPU acceleration. BiGMT consistently outperforms ExTreeM, PPP, and FTM-Tree on RTX4090 and H200, achieving speedups of up to 23.95× and demonstrating superior parallel efficiency and scalability.

Cross-view geo-localization aims at determining the geographic location of a query image by matching the reference images. The matching pairs can be captured from diverse perspectives, such as those from satellites and drones. Most existing methods are supervised that require input of location-labeled images or matched and unmatched image pairs for training, resulting in high labor costs. Moreover, current unsupervised methods perform instances matching directly between different perspectives with dramatic discrepancies, resulting in poor performance. To address these issues, this paper proposes a novel matching and alignment framework from coarse instance-cluster level to fine intermediate instance level for unsupervised cross-view geo-localization. We first introduces cluster-based contrastive learning, assigning pseudo-labels to the instances and generate clusters within each view. Then we design a cross-view location alignment module that fully exploits the feature relationships between instances and clusters for intra- and inter-views. Finally, we design an intermediate state transition module that facilitates further alignment between views by constructing intermediate states and bringing both views closer to the intermediate domain simultaneously. Extensive experiments demonstrate that our method surpasses state-of-the-art unsupervised cross-view geo-localization methods and even achieves comparable performance to state-of-the-art supervised methods.

Machine learning has demonstrated transformative potential for database operations, such as query optimization and in-database data analytics. However, dynamic database environments, characterized by frequent updates and evolving data distributions, introduce concept drift, which leads to performance degradation for learned models and limits their practical applicability. Addressing this challenge requires efficient frameworks capable of adapting to shifting concepts while minimizing the overhead of retraining or fine-tuning. In this paper, we propose FLAIR, an online adaptation framework that introduces a new paradigm called \textit{in-context adaptation} for learned database operations. FLAIR leverages the inherent property of data systems, i.e., immediate availability of execution results for predictions, to enable dynamic context construction. By formalizing adaptation as $f:(\mathbf{x} | \mathcal{C}_t) \to \mathbf{y}$, with $\mathcal{C}_t$ representing a dynamic context memory, FLAIR delivers predictions aligned with the current concept, eliminating the need for runtime parameter optimization. To achieve this, FLAIR integrates two key modules: a Task Featurization Module for encoding task-specific features into standardized representations, and a Dynamic Decision Engine, pre-trained via Bayesian meta-training, to adapt seamlessly using contextual information at runtime. Extensive experiments across key database tasks demonstrate that FLAIR outperforms state-of-the-art baselines, achieving up to $5.2\times$ faster adaptation and reducing error by 22.5\% for cardinality estimation.

In this paper, we address the cooperative robust parallel operation problem of an electric drive shaft system. In contrast to prior research, this work explicitly incorporates system uncertainties and external disturbances affecting both the shaft and the motors, enhancing the robustness and practicality of the proposed approach. To address this challenge, we first establish a dynamic output feedback controller utilizing the internal model principle. Then, we demonstrate that the cooperative robust parallel operation of the electric drive shaft system can be achieved under directed communication networks, effectively overcoming the adverse effects of system uncertainties and external disturbances. Finally, the efficacy of our proposed distributed controller is rigorously validated through its application to an electric drive shaft system equipped with five actuator motors.

This paper presents PartCAD, a novel framework for reconstructing CAD modeling sequences directly from point clouds by projection-guided, part-aware geometry reasoning. It consists of (1) an autoregressive approach that decomposes point clouds into part-aware latent representations, serving as interpretable anchors for CAD generation; (2) a projection guidance module that provides explicit cues about underlying design intent via triplane projections; and (3) a non-autoregressive decoder to generate sketch-extrusion parameters in a single forward pass, enabling efficient and structurally coherent CAD instruction synthesis. By bridging geometric signals and semantic understanding, PartCAD tackles the challenge of reconstructing editable CAD models—capturing underlying design processes—from 3D point clouds. Extensive experiments show that PartCAD significantly outperforms existing methods for CAD instruction generation in both accuracy and robustness. The work sheds light on part-driven reconstruction of interpretable CAD models, opening new avenues in reverse engineering and CAD automation.

With the continuous improvement of living standards, the tourism industry is also constantly developing. In order to meet the needs of most tourists, relevant management personnel have proposed tourism path planning methods. However, currently, many path planning methods cannot accurately analyze multi-dimensional portraits of tourists, and there are also issues with weak rationality and low tourist satisfaction in tourism path planning. Therefore, this study uses convolutional attention mechanism to optimize the user multi-dimensional portrait scenario recommendation algorithm based on support vector machine and collaborative filtering algorithm. A tourism path planning method is designed for this optimization algorithm. The research first tested the optimization algorithm. The coverage of this optimization algorithm was the widest, with an error rate of only 0.23%, and a recommendation time of only 2.3ms, significantly better than comparison algorithms. The path planning method was analyzed. The path planning accuracy reached 97.1%, the planning time was only 1.2s, and the rationality of the planned tourist route reached 0.95. The satisfaction rate with this route was 97.4%. From the above results, it can be seen that the proposed tourism path planning method has a certain effect on improving tourists’ satisfaction with the tourism route and has a positive impact on the development of the tourism industry.

The flexible job shop scheduling problem for processing machines and transportation vehicles (FJSP_PT) has garnered significant attention from academia and industry. Due to the inclusion of transportation vehicle scheduling in the scheduling problem of flexible manufacturing systems, solving FJSP_PT becomes more challenging and significantly more practically relevant compared to the flexible job shop scheduling problem. We summarize the assumptions, constraints, objective functions, and benchmarks of FJSP_PT. Then, statistical analysis is conducted on the literature up to 2023, including journals, number of articles published each year, and solution algorithms. We analyze recent literature on FJSP_PT, categorizing it based on algorithms into exact algorithms, heuristic algorithms, meta-heuristic algorithms, and swarm intelligence based algorithms. Finally, the research trends and challenges faced by FJSP_PT are summarized.

In this paper, a new Messenger Routing Problem (MRP) is studied, which is motivated by Unmanned Aerial Vehicles (UAVs) accessing Unmanned Ground Vehicles (UGVs) to deliver information in UAV-UGV coordination systems. The objective is to minimize the longest path among multiple messengers, ensuring fast and reliable information transmission. Two key challenges arise in tackling this problem. First, the targets are moving, incurring the travel cost between targets to vary with the travel process. Second, the messengers accessing the neighborhood of targets needs to satisfy the communication time constraint. Based on the idea of decoupling, a two-phase planner is proposed to sequentially determine global access sequence and optimize local access locations. In the first phase, a motion prediction module is introduced in the Adaptive Large Neighborhood Search (ALNS) framework to deal with the dynamic characteristics of MRP. In the second phase, an efficient bisection sampling method based on the prediction points is proposed to obtain a shorter access path while satisfying the communication time constraint. Finally, the effectiveness and efficiency of the proposed method are demonstrated by performance evaluation and comparison with the state-of-the-art algorithms.

3D human pose estimation (3DHPE) from a single monocular RGB image is fundamental in many image-related fields, such as virtual reality, motion analysis, and human-computer interaction. To improve estimation accuracy, existing works typically integrate complex networks or divide monocular 3DHPE into multiple stages. However, complicating the estimation process to improve the estimation accuracy sacrifices the estimation speed and limits its application. To alleviate this, we propose AITEPose, an end-to-end model, which achieves higher monocular 3DHPE accuracy with a simpler model structure. Specifically, inspired by online knowledge distillation, we design an Auxiliary-Information-Driven Training Enhancement (AITE) framework. In the AITE framework, during training, an adjustment network is introduced between the prediction network and the loss function to incorporate auxiliary information and enhance the training process. Notably, the adjustment network is constructed by developing a novel cascaded Disturbance-Correction Module (DCM). It adjusts the poses to get more accurate results based on ground-truth bone lengths. Both AITE and DCM are employed only during training, thereby improving training outcomes without complicating the inference process. The AITEPose model achieves state-of-the-art performance for single-frame monocular 3DHPE on the most comprehensive dataset Human3.6M. To further validate the effectiveness of AITE and DCM, we design a monocular 2DHPE model, AITEPose2D, and conduct extensive ablation experiments on the COCO2017 dataset, demonstrating the robustness and generalizability of our proposed AITEPose.

This article investigates stochastic differential games involving multiple attackers, defenders, and a single target, with their interactions defined by a distributed topology. By leveraging principles of topological graph theory, a distributed design strategy is developed that operates without requiring global information, thereby minimizing system coupling. Additionally, this study extends the analysis to incorporate stochastic elements into the target-attackers-defenders games, moving beyond the scope of deterministic differential games. Using the direct method of completing the square and the Radon-Nikodym derivative, we derive optimal distributed control strategies for two scenarios: one where the target follows a predefined trajectory and another where it has free maneuverability. In both scenarios, our research demonstrates the effectiveness of the designed control strategies in driving the system toward a Nash equilibrium. Notably, our algorithm eliminates the need to solve the coupled Hamilton-Jacobi equation, significantly reducing computational complexity. To validate the effectiveness of the proposed control strategies, numerical simulations are presented in this article.

Robotic grasping is a crucial topic in robotics and computer vision, with broad applications in industrial production and intelligent manufacturing. Although some methods have begun addressing instance-level grasping, most remain limited to predefined instances and categories, lacking flexibility for open-vocabulary grasp prediction based on user-specified instructions. To address this, we propose RoG-SAM, a language-driven, instance-level grasp detection framework built on Segment Anything Model (SAM). RoG-SAM utilizes open-vocabulary prompts for object localization and grasp pose prediction, adapting SAM through transfer learning with encoder adapters and multi-head decoders to extend its segmentation capabilities to grasp pose estimation. Experimental results show that RoG-SAM achieves competitive performance on single-object datasets (Cornell and Jacquard) and cluttered datasets (GraspNet-1Billion and OCID), with instance-level accuracies of 91.2% and 90.1%, respectively, while using only 28.3% of SAM's trainable parameters. The effectiveness of RoG-SAM was also validated in real-world environments. A demonstration video is available athttps://www.youtube.com/playlist?list=PL7et4nGJAImLGytsJbglGbXl1hacA2dy_.

This article investigates the distributed predefined-time (PT) fault-tolerant control for nonlinear multiagent systems (NNMSs) with nonaffine faults. A novel distributed PT control scheme is proposed based on a new PT theorem. The switching functions are introduced into distributed control signals to avoid the singularity problem. The second-order filter technology is designed to solve “explosion of complexity” issues. The newly designed PT compensation signals are utilized to eliminate filter errors. The proposed distributed PT controller based on local information can guarantee that all signals of the closed-loop system are PT bounded, and consensus errors of NNMSs can be enforced to a small region around the origin within a predefined time. Compared with the finite/fixed time, the predefined time is an exact value given in advance, which remains constant regardless of initial states and control parameters. Finally, two comparative simulations are provided to demonstrate the presented strategy.

Facility location is a crucial aspect of spatial optimization with broad applications in urban planning. Specifically, the multi-period problem involves spatial and temporal information, making it challenging to solve. Existing research mainly focuses on heuristic methods, which depend on complex hand-crafted techniques. In this paper, we propose a novel method based on Deep Reinforcement Learning (DRL) to solve pk-median Dynamic Location Problem (DLP-pk). Different from classical heuristic methods, our method avoids intricate designs and considers the temporal impacts of decisions. We are the first to apply DRL to the multi-period facility location problem. Our method adopts the encoder-decoder architecture and utilizes a specialized structure to capture the temporal features across different periods. On the one hand, we introduce the Gated Recurrent Units (GRU) to encode temporal information, including dynamic coverage and dynamic costs. On the other hand, we design an attention-based decoder that allows the model to capture long-term dependencies in decision-making. Experimental results from small-size to large-size demonstrate that our method can quickly provide high-quality solutions without relying heavily on expert knowledge, offering opportunities for efficiently solving multi-period facility location problems. Additionally, our method is up to two orders of magnitude faster than the exact solver Gurobi and demonstrates great generalization abilities.

Most RGB-D-based SLAM methods assume texture-rich environments, making them susceptible to significant tracking errors or complete failures in the absence of texture features. Moreover, many existing methods encounter substantial rotation estimation errors, leading to long-term drift in tracking. This paper proposes a novel structured-constrained RGB-D SLAM method (STL-SLAM) for texture-limited environments. Compared to the existing methods, STL-SLAM can deal with environments without abundant texture information and significantly reduce long-term drift caused by rotation estimation errors. We assess the distribution complexity of pixels in an image by calculating the information entropy and pre-processing accordingly. We also present an efficient Manhattan Frames (MF) detection strategy based on orthogonal planes and lines. If MF is detected, we decouple rotation and translation, estimate drift-free rotation based on the Manhattan World (MW) coordinate system, and then estimate translation by minimizing the re-projection error of point, line, and plane features. In non-Manhattan Frames, the 6-DoF pose estimation is performed holistically, with the incorporation of structural constraints of parallel and perpendicular planes, as well as parallel and vertical lines, into the optimization process. Finally, we evaluate our method on public datasets and in real-world environments, which shows that our proposed method achieves superior performance compared to its counterparts.

Large vision-language models (LVLMs) are markedly proficient in deriving visual representations guided by natural language. Recent explorations have utilized LVLMs to tackle zero-shot visual anomaly detection (VAD) challenges by pairing images with textual descriptions indicative of normal and abnormal conditions, referred to as anomaly prompts. However, existing approaches depend on static anomaly prompts that are prone to cross-semantic ambiguity, and prioritize global image-level representations over crucial local pixel-level image-to-text alignment that is necessary for accurate anomaly localization. In this paper, we present ALFA, a training-free approach designed to address these challenges via a unified model. We propose a run-time prompt adaptation strategy, which first generates informative anomaly prompts to leverage the capabilities of a large language model (LLM). This strategy is enhanced by a contextual scoring mechanism for per-image anomaly prompt adaptation and cross-semantic ambiguity mitigation. We further introduce a novel fine-grained aligner to fuse local pixel-level semantics for precise anomaly localization, by projecting the image-text alignment from global to local semantic spaces. Extensive evaluations on the challenging MVTec and VisA datasets confirm ALFA's effectiveness in harnessing the language potential for zero-shot VAD, achieving significant PRO improvements of 12.1% on MVTec AD and 8.9% on VisA compared to state-of-the-art zero-shot VAD approaches.

Implanted devices are increasingly used in chronic disease monitoring, but face challenges in energy autonomy. This article presents a novel wireless power transfer (WPT) method for self-sustained medical implants using Halbach array-based magnetic plucking and piezoelectric transduction. The wearable-implantable coupled system consists of a piezoelectric receiver within the implant to receive power and a near-field magnetic power transmitter as a wearable device. To deliver power over greater distances through the human body, the transmitter features a rotating magnetic Halbach array powered by a miniature motor, or by human motion, to generate an alternating magnetic field. The use of low-frequency rotating magnetic fields periodically excites a cantilevered piezoelectric beam with a tip magnet to realize WPT. A theoretical model that includes magnetic coupling, piezoelectric transduction and receiver beam dynamics has been established to study the electro-magneto-mechanical dynamics of this WPT system. The effectiveness of the Halbach array for extended power transfer is examined through theoretical modeling and numerical simulation, showing a 37.2% enhancement of the magnetic forces. A prototype was also fabricated and tested to examine the WPT performance. The established wireless power link can provide sufficient power ($\sim 32~\mu $W) over a large transmission distance (22 mm), providing a potential battery-free solution for the self-sustained Internet of Implanted Things (IoIT) for personalized healthcare.

The flexible job-shop co-scheduling problem (FJCSP) for processing machines and automated guided vehicles (AGVs) in a flexible manufacturing system (FMS) has attracted more attention with the aim of improving production efficiency. In FMS, AGVs in charge of transporting jobs realize the flexible linkage of operations between different processing machines. The added interdependence between transporting and processing tasks brings more difficulties than the traditional flexible job-shop scheduling problem (FJSP). In this paper, the mathematical model of FJCSP is formulated to minimize the makespan. Considering the feature similarity of FJCSP with FJSP and AGV-routing problem in different cases, a multi-view modeling-based hybrid algorithm consisting of an estimation of distribution algorithm (EDA) and an ant colony optimization (ACO) is proposed. In EDA, a probability model abstracts the information in superior solutions about the operation sequencing and the rule selection for scheduling machines and AGVs. In ACO, a job-path pheromone model and an AGV-path pheromone model are designed to jointly select the job-machine-AGV combination with shorter processing time and transportation time. In the proposed hybrid algorithm, EDA and ACO generate solutions independently and achieve cooperation by sharing elites. An adaptive parameter is designed to regulate the use of the two methods to adapt to the varying demands of multi-view modeling in different cases and search stages. Furthermore, a local search with a three-layer operator based on the critical path method is proposed to balance exploration and exploitation in solution space. Finally, computational experiments involving a case study verified the advantage of the multi-view modeling-based hybrid algorithm in comparison with the state-of-the-art approaches.Note to Practitioners—This paper was motivated by the optimization problem of scheduling machines and automated guided vehicles (AGVs) in flexible manufacturing system (FMS). In FMS with AGVs, the transportation stages for jobs by AGVs significantly impact the overall production efficiency of the FMS and cannot be overlooked. This paper suggested a hybrid evolutionary algorithm using an estimation of distribution algorithm (EDA), an ant colony optimization (ACO) and a local search algorithm based on the critical path method. In the proposed hybrid algorithm, an adaptive parameter is introduced to regulate the utilization of EDA and ACO in generating a new population. This paper presents a mathematical characterization of the scheduling problem and subsequently outlines the step-by-step design of the hybrid algorithm. Computational experiments, including a case study, demonstrate that the hybrid algorithm exhibits adaptability to various instances and outperforms state-of-the-art approaches.

This article investigates the cooperative control of complex nonlinear multiagent systems (CNMASs), in which the agents suffer from nonaffine faults and the control directions of some agents are unknown. A finite-time adaptive control scheme is presented for the CNMASs. A finite-time command filter is designed to solve the "explosion of complexity" issues, overcome chattering issues, and relax the limitations of the filter input. The impact of filter errors is alleviated by an improved error compensation mechanism. Based on piecewise Nussbaum functions, the partially unknown control direction is addressed. The proposed finite-time cooperative control strategy on the basis of local information can ensure that all signals in the closed-loop system are finite-time bounded, and the absolute value of the cooperative control errors can converge to a given upper bound in a finite time. The rapidity and robustness of the proposed method are verified by two comparative simulation examples. A real multirobot cooperative control experiment is used to verify the effectiveness of the presented method.

This article studies the tracking control of a class of nonlinear systems with input saturation, subject to nonaffine faults and unknown control direction. A fault-tolerant command filtered control (CFC) method based on adaptive neural networks (NNs) is proposed for this kind of nonlinear system. First, the combination of CFC and error compensation overcomes the "explosion of complexity" issue and alleviates the impact of filter errors. Then, a set of radial basis function NNs is constructed to approximate the unknown nonlinear items containing the nonaffine fault function. Additionally, the issue of unknown control direction in the system is effectively resolved by using Nussbaum gain technology. It is proven that the designed controller can ensure that all signals in the closed-loop system are bounded and convergent, and the upper bound of the absolute value of system tracking error is given. Finally, three comparative simulation results are illustrated to show the effectiveness of the proposed method.

Flexible manufacturing has given rise to complex scheduling problems such as the flexible job shop scheduling problem (FJSP). In FJSP, operations can be processed on multiple machines, leading to intricate relationships between operations and machines. Recent works have employed deep reinforcement learning (DRL) to learn priority dispatching rules (PDRs) for solving FJSP. However, the quality of solutions still has room for improvement relative to that by the exact methods such as OR-Tools. To address this issue, this article presents a novel end-to-end learning framework that weds the merits of self-attention models for deep feature extraction and DRL for scalable decision-making. The complex relationships between operations and machines are represented precisely and concisely, for which a dual-attention network (DAN) comprising several interconnected operation message attention blocks and machine message attention blocks is proposed. The DAN exploits the complicated relationships to construct production-adaptive operation and machine features to support high-quality decision-making. Experimental results using synthetic data as well as public benchmarks corroborate that the proposed approach outperforms both traditional PDRs and the state-of-the-art DRL method. Moreover, it achieves results comparable to exact methods in certain cases and demonstrates favorable generalization ability to large-scale and real-world unseen FJSP tasks.

Reidentification (Re-ID) is a crucial computer vision application with a variety of potential uses in many maritime scenarios, including search, rescue, and surveillance. However, the development of advanced boat reidentification (Boat Re-ID) algorithms necessitates the availability of large-scale Re-ID datasets for model training and evaluation. Inspired by scenarios engineering, this study proposes a new framework for automatically generating a realistic synthetic dataset for boat Re-ID investigation. The synthetic dataset contains 107 boat models and various visual conditions in 36 real backgrounds. The use of synthetic datasets enables the learning-based Re-ID algorithm’s performance to be quantitatively verificated under varying imaging conditions. Nonetheless, our experiments prove that synthetic datasets are inadequate to handle real-world challenges. Therefore, we present a domain adaptation approach that integrates both real and synthetic data to create trustworthy models. This approach employs a multistep training strategy, gradient reversal layer and novel loss functions to preserve the features from two distribution dataset domains. The results of the experiments demonstrate that 1) synthetic datasets can be employed to train boat Re-ID algorithms and quantitatively test the performance of these algorithms under diverse imaging conditions and 2) our approach utilizes the attributes of the two data domains (real and synthetic) to achieve exceptional performance in real-world applications.

The emergence of a robotic mobile fulfillment system (RMFS) provides an automated solution for e-commerce warehousing to improve productivity and reduce labor costs. This article studies the order picking optimization problem in RMFS, which simultaneously decides the assignment of orders and racks to multiple picking stations. Although this problem has been widely studied in recent years, it is still very challenging for existing methods to solve large-scale instances effectively (e.g., more than 200 orders and 500 racks). To overcome this difficulty to meet the real-world needs, we propose a fulfillment-focused simultaneous assignment (FFSA) method. The proposed FFSA comprises two stages: 1) compression and 2) simultaneous assignment. The compression stage employs a hybrid adaptive large neighborhood search (ALNS) strategy to establish a reduced set of critical racks that can fulfill the demand of all orders. In the simultaneous assignment stage, we develop a marginal-return-based assignment with candidate strategy (MRACS) to simultaneously assign orders and critical racks to picking stations. MRACS takes into account three fulfillment-focused measurements to depict the product supply relationship between the demand of orders and the inventory on critical racks. These measurements are further integrated into the effective heuristics with sufficient problem-specific knowledge to obtain a high-quality solution. Experimental results show that our method significantly outperforms representative algorithms on both synthetic data and large-scale real-world data.

This paper presents Triangulation Residual loss (TR loss) for multiview 3D pose estimation in a data-efficient manner. Existing 3D supervised models usually require large-scale 3D annotated datasets, but the amount of existing data is still insufficient to train supervised models to achieve ideal performance, especially for animal pose estimation. To employ unlabeled multiview data for training, previous epipolar-based consistency provides a self-supervised loss that considers only the local consistency in pairwise views, resulting in limited performance and heavy calculations. In contrast, TR loss enables self-supervision with global multiview geometric consistency. Starting from initial 2D keypoint estimates, the TR loss can fine-tune the corresponding 2D detector without 3D supervision by simply minimizing the smallest singular value of the triangulation matrix in an end-to-end fashion. Our method achieves the state-of-the-art 25.8mm MPJPE and competitive 28.7mm MPJPE with only 5\% 2D labeled training data on the Human3.6M dataset. Experiments on animals such as mice demonstrate our TR loss's data-efficient training ability.

Real-time analytics and decision-making require online anomaly detection (OAD) to handle drifts in data streams efficiently and effectively. Unfortunately, existing approaches are often constrained by their limited detection capacity and slow adaptation to evolving data streams, inhibiting their efficacy and efficiency in handling concept drift , which is a major challenge in evolving data streams. In this paper, we introduce METER, a novel dynamic concept adaptation framework that introduces a new paradigm for OAD. METER addresses concept drift by first training a base detection model on historical data to capture recurring central concepts , and then learning to dynamically adapt to new concepts in data streams upon detecting concept drift. Particularly, METER employs a novel dynamic concept adaptation technique that leverages a hypernetwork to dynamically generate the parameter shift of the base detection model, providing a more effective and efficient solution than conventional retraining or fine-tuning approaches. Further, METER incorporates a lightweight drift detection controller, underpinned by evidential deep learning, to support robust and interpretable concept drift detection. We conduct an extensive experimental evaluation, and the results show that METER significantly outperforms existing OAD approaches in various application scenarios.

Unsupervised anomaly detection (AD) is a challenging problem in the data mining community. Clustering-based AD methods aim to group normal data points into clusters and then regard a point belonging to none of the clusters as an anomaly. However, they may suffer from the problems of unknown cluster numbers and arbitrary cluster shapes. This paper presents a novel clustering-based AD method named Density-increasing Path (DIP) to tackle these challenges. DIP searches a path for each data point. The path starts at the data point itself, passes through several points with monotonically increasing densities, and ends at a density peak. Further, DIP defines the climbing difficulty of each path by combining the distance and density increment of each step along the path, which can be regarded as the anomaly score of the path starting point. DIP can adaptively decide the number of peaks to address the challenge of unknown cluster numbers. Since DIP requires the path to pass several points rather than directly reaching the peak, it handles arbitrary cluster shapes. We also propose the ensemble DIP to improve prediction accuracy. The experimental results on four synthetic datasets and eleven real-world benchmarks demonstrate that DIP outperforms existing methods.

Most anchor-free methods perform object detection using dense recommendation, which assumes that one point can simultaneously conduct accurate category prediction and regression estimation. However, due to different task drivers, valid features for classification and regression may locate at distinct areas in the training phase. This problem is called feature misalignment. To solve it, we propose a new feature alignment method based on anchor-free object detector. Firstly, a global receptive field adaptor (G-RFA) is designed by incorporating the feature pyramid networks (FPN) with the global attention mechanism, and forward features are further fine-tuned with a deformable-subnet (De-Subnet) to remove the influence of redundant contextual information. Then, a new feature filter strategy with a misalignment score is proposed to guide the network to focus on sampling points with aligned features. In addition, we establish mutually independent multi-layer quality distributions to model the priori information of an object on different FPN levels. Equipped with our method, the classification and regression features are aligned, and the generated foreground weight map converges to the centers of classification and regression heatmaps. Experimental results show that without bells and whistles, our method achieves 49.3% AP on MS COCO test-dev under the default 2x training schedule, outperforming related methods. Besides, experiments on PASCAL VOC demonstrate the generalization ability of our method. Code is available at https://github.com/GFENGG/featurealign.

In this article, the robust output regulation problem of the linear uncertain system is investigated by the event-triggered control approach. Recently, the same problem is addressed by an event-triggered control law where the Zeno behavior may happen when time tends to infinity. In comparison, a class of event-triggered control laws is developed to achieve output regulation exactly, and meanwhile, explicitly exclude the Zeno behavior for all time. In particular, a dynamic triggering mechanism is first developed by introducing a dynamic changing variable with specific dynamics. Then, by the internal model principle, a class of dynamic output feedback control laws is designed. Later, a rigorous proof is provided to show that the tracking error of the system converges to zero asymptotically while prohibiting the Zeno behavior for all time. Finally, we give an example to illustrate our control approach.

This article studies the fixed-time consensus tracking control of nonlinear multiagent systems (NNMASs) suffering from nonaffine faults. A fixed-time command filter is constructed to solve the “explosion of complexity” issue, and a novel fixed-time error compensation mechanism is designed to eliminate filtering errors. The adaptive fuzzy control technique is introduced to deal with the unknown nonlinear terms containing nonaffine fault functions. A new fixed-time fault-tolerant control scheme based on the modified prescribed performance technology is proposed for the NNMASs, which ensures that the closed-loop system satisfies the practical fixed-time stability. In addition, the consensus tracking errors of the NNMASs converge to a given range within a prescribed performance bound in a fixed time. Finally, comparative simulation results show the effectiveness of the proposed method.

Virtual simulated scenes are becoming a critical part of autonomous driving. In the context of knowledge automation and machine learning, simulated images are widely used for visual environmental perception. However, even the most inspirational applications have not fully exploited the potential of simulated images in solving real-world problems. In this article, we propose a novel framework “ParallelEye Pipeline,” which uses image-to-image translation and simulated images to automatically generate realistic synthetic images with multiple ground-truth annotations. Specifically, this method has three steps: first, we use Unity3D software to simulate driving scenarios and generate simulated image pairs (including raw images and six ground-truth labels) from the simulated scenes; second, advanced image-to-image translation algorithms can generate realistic and high-resolution synthetic images from simulated image pairs; third, we exploit publicly datasets, simulated images, and synthetic images to conduct experiments for visual perception. The experimental results suggest: 1) synthetic images and simulated images can improve the performance of detectors in real autonomous driving scenarios and 2) image-to-image translation algorithms can be affected by occlusion condition.

The unmanned air-ground vehicle system has been successfully applied in civil and military domains. Collaborative vision-based target geolocation with this system can provide an enduring and accurate estimate of moving target state. Traditional decentralized information filter (DIF) treated each platform in the system identically. In fact, the observation capabilities of aerial and ground platform typically differ from each other due to different configurations and changing sensor noises. Without considering these differences, the resources of each platform cannot be fully utilized. To handle the issue, we develop a weighted DIF for geolocating of moving targets via air-ground collaboration. Specifically, it can produce a weighted factor autonomously for each platform based on the similarity of tracks from the air-ground system. Then, it is able to have more accurate global estimates than the traditional filter. Finally, simulation experiments and actual tests are conducted and the results are presented to validate the efficacy of the proposed method. Additional details can be seen in our video submission.

The problem of order and rack assignment to picking stations (ORAPS) is a key joint optimization problem in the order picking process of the robotic mobile fulfillment system (RMFS). Given a set of customer orders and a set of mobile racks, the goal of this problem is to simultaneously assign orders and racks to multiple picking stations so that the racks can provide products to meet the demand of orders by the minimal number of visits to all picking stations. In this article, we build a mathematical model for the ORAPS, which considers all orders assignment with the allocable capacity interval of the picking station and the rack product capacity limitation. To solve the ORAPS problem, we propose a two-stage hybrid heuristic algorithm (TS-HHA) consisting of the reducing stage and the assigning stage. In the reducing stage, a scheme of dynamic programming (DP) is introduced to find a critical rack set, which can focus attention on the most promising racks and increase the speed of problem-solving. In the assigning stage, we propose an optimization strategy that combines a constructive heuristic algorithm and adaptive neighborhood search (CH-ANS). It can generate a high-quality simultaneous assignment scheme and further improve its quality effectively. The computational results show that our proposed algorithm performs better than its competitors on both simulation and practical instances of the ORAPS problem. Note to Practitioners—This article investigates the order and rack assignment to picking stations (ORAPS) problem in the robotic mobile fulfillment system (RMFS) originated from a renowned Chinese logistics platform. We build a more comprehensive and reasonable mathematical programming model for this ORAPS problem and propose a two-stage hybrid heuristic algorithm (TS-HHA) strategy, which can simultaneously assign orders and racks to multiple picking stations so that all customer orders can be processed. The experimental results show that our TS-HHA performs at least 35% better than its competitors on all simulation and practical instances set. We believe that this research work can serve as a kind of generic framework for practical ORAPS problems and be very helpful for operating the RMFS efficiently.

This article focuses on event-triggered consensus control for multiagent systems subject to sensor faults or noises. First, a descriptor state observer with a low-pass filtering characteristic being developed for each agent using output information. The convergence regions of estimation errors can be reduced by a nonsingular suppression matrix. Leader-follower event-triggered consensus protocols with continuous-time communication are designed for multiagent systems based on the estimated states. By virtue of the Jordan form of the Laplacian matrix, the stability conditions are derived by using the Lyapunov analysis. Then, new self-triggered consensus protocols are designed for the multiagent systems to remove the requirement of the continuous monitoring triggering condition and continuous communication simultaneously. The triggering interval is proved greater than 0, and the Zeno behavior is excluded for all agents. Finally, numerical simulations are conducted to demonstrate the effectiveness of the proposed design.

Unsupervised domain adaptation (DA) aims to perform classification tasks on the target domain by leveraging rich labeled data in the existing source domain. The key insight of DA is to reduce domain divergence by learning domain-invariant features or transferable instances. Despite its rapid development, there still exist several challenges to explore. At the feature level, aligning both domains only in a single way (i.e., geometrical or statistical) has limited ability to reduce the domain divergence. At the instance level, interfering instances often obstruct learning a discriminant subspace when performing the geometrical alignment. At the classifier level, only minimizing the empirical risk on the source domain may result in a negative transfer. To tackle these challenges, this article proposes a novel DA method, called discriminant geometrical and statistical alignment (DGSA). DGSA first aligns the geometrical structure of both domains by projecting original space into a Grassmann manifold, then matches the statistical distributions of both domains by minimizing their maximum mean discrepancy on the manifold. In the former step, DGSA only selects the density peaks to learn the Grassmann manifold and so to reduce the influences of interfering instances. In addition, DGSA exploits the high-confidence soft labels of target landmarks to learn a more discriminant manifold. In the latter step, a structural risk minimization (SRM) classifier is learned to match the distributions (both marginal and conditional) and predict the target labels at the same time. Extensive experiments on objection recognition and human activity recognition tasks demonstrate that DGSA can achieve better performance than the comparison methods.

Anomaly detection (AD) has attracted great interest in the data mining community. With the development of deep learning, various deep autoencoders have been used and modified to solve AD problems due to their efficient data coding and reconstruction mechanisms. However, such methods still suffer challenges when solving some practical AD tasks. On the one hand, an AD dataset may contain diverse normal patterns rather than a universal pattern. Specifically, the normal data usually distribute in multiple clusters; meanwhile, the exact number of clusters is hard to know in practice. On the other hand, most existing autoencoder-based methods focus on encoding normal features but have not considered exploring the characteristics of abnormal data. To tackle these challenges, this article proposes a novel autoencoder-based AD model, the attract-repel encoder (ARE). ARE selects some landmarks in the encoding space to represent the diverse normal patterns. Besides, ARE can adaptively update the landmarks and their quantity during training. Then this article proposes the attract-repel loss (AR loss) function to train ARE. AR loss attracts normal samples to landmarks and repels anomalies away from landmarks so that it can learn both normal and abnormal features. Finally, ARE computes a sample's anomaly score by summing up its reconstruction error and its distance to the landmarks. Moreover, ARE can be trained either semisupervised or unsupervised. This article presents comprehensive experiments to evaluate the effectiveness of our approach.

Gene expression profiling has played a significant role in the identification and classification of tumor molecules. In gene expression data, only a few feature genes are closely related to tumors. It is a challenging task to select highly discriminative feature genes, and existing methods fail to deal with this problem efficiently. This article proposes a novel metaheuristic approach for gene feature extraction, called variable neighborhood learning Harris Hawks optimizer (VNLHHO). First, the F-score is used for a primary selection of the genes in gene expression data to narrow down the selection range of the feature genes. Subsequently, a variable neighborhood learning strategy is constructed to balance the global exploration and local exploitation of the Harris Hawks optimization. Finally, mutation operations are employed to increase the diversity of the population, so as to prevent the algorithm from falling into a local optimum. In addition, a novel activation function is used to convert the continuous solution of the VNLHHO into binary values, and a naive Bayesian classifier is utilized as a fitness function to select feature genes that can help classify biological tissues of binary and multi-class cancers. An experiment is conducted on gene expression profile data of eight types of tumors. The results show that the classification accuracy of the VNLHHO is greater than 96.128% for tumors in the colon, nervous system and lungs and 100% for the rest. We compare seven other algorithms and demonstrate the superiority of the VNLHHO in terms of the classification accuracy, fitness value and AUC value in feature selection for gene expression data.

Sensor-based human activity recognition (HAR) aims to recognize a human's physical actions by using sensors attached to different body parts. As a user-specific application, HAR often suffers poor generalization from training on an individual to testing on another individual, or from one body part to another body part. To tackle this cross-domain HAR problem, this article proposes a domain adaptation (DA) method called local domain adaptation (LDA), whose core is to align cluster-to-cluster distributions between the source domain and the target domain. On the one hand, LDA differs from existing set-to-set alignment by reducing the distribution discrepancy at a finer granularity. On the other hand, LDA is superior to the class-to-class alignment because it can provide more accurate soft labels for the target domain. Specifically, LDA contains three main steps: 1) groups the activity class into several high-level abstract clusters; 2) maps the original data of each cluster in both domains into the same low-dimension subspace to align the intracluster data distribution; 3) predicts the class labels for target domain in the low-dimension subspace. Experimental results on two public HAR benchmark datasets show that LDA outperforms state-of-the-art DA methods for the cross-domain HAR.

Traditional monocular vision localization methods are usually suitable for short-range area and indoor relative positioning tasks. This paper presents MGG, a novel monocular global geolocation method for outdoor long-range targets. This method takes a single RGB image combined with necessary navigation parameters as input and outputs targets' GPS information under the Global Navigation Satellite System (GNSS). In MGG, we first design a camera pose correction method via pixel mapping to correct the pose of the camera. Then, we use anchor-based methods to improve the detection ability for long-range targets with small image regions. Next, the local monocular vision model (LMVM) with a local structure coefficient is proposed to establish an accurate 2D-to-3D mapping relationship. Subsequently, a soft correspondence constraint (SCC) is presented to solve the local structure coefficient, which can weaken the coupling degree between detection and localization. Finally, targets can be geolocated through optimization theory-based methods and a series of coordinate transformations. Furthermore, we demonstrate the importance of focal length on solving the error explosion problem in locating long-range targets with monocular vision. Extensive experiments on the challenging KITTI dataset as well as applications in outdoor environments with targets located at a long range of up to 150 meters show the superiority of our method.

Unsupervised anomaly detection for time series is of great importance for various applications, such as Web monitoring, medical monitoring, and device fault diagnosis. Time series anomaly detection (TSAD) aims to find the observations that most different from others in a sequence of observations. With the development of deep learning, deep-autoencoder-based methods achieve state-of-the-art performance. These methods are usually able to find single anomaly points but fail to detect the anomaly segment and the change point. To tackle this problem, this paper proposes a novel TSAD method, which consists of a bidirectional LSTM (BiLSTM) autoencoder and a subsequent Gaussian segmentation model. BiLSTM encodes a time series in a predictive format from both positive and negative time directions, then outputs the latent feature vectors and restructured errors. After that, the latent features are used to find anomaly segments by the Gaussian segment model; the restructured errors are used to find change points and extreme single anomaly by a scoring function. In this way, our method can find all three kinds of anomaly points. Experiments on two real-world datasets demonstrate the effectiveness of the proposed method.

Binocular vision is a passive method of simulating the human visual principle to perceive the distance to a target. Traditional binocular vision applied to target localization is usually suitable for short-range area and indoor environment. This paper presents a novel vision-based geolocation method for long-range targets in outdoor environment, using handheld electronic devices such as smart phones and tablets. This method solves the problems in long-range localization and determining geographic coordinates of the targets in outdoor environment. It is noted that these sensors necessary for binocular vision geolocation such as the camera, GPS, and inertial measurement unit (IMU), are intergrated in these handheld electronic devices. This method, employing binocular localization model and coordinate transformations, is provided for these handheld electronic devices to obtain the GPS coordinates of the targets. Finally, two types of handheld electronic devices are used to conduct the experiments for targets in long range up to 500m. The experimental results show that this method yields the target geolocation accuracy along horizontal direction with nearly 20m, achieving comparable or even better performance than monocular vision methods.

Despite the significant progress in end-to-end (E2E) automatic speech recognition (ASR), E2E ASR for low resourced code-switching (CS) speech has not been well studied. In this work, we describe an E2E ASR pipeline for the recognition of CS speech in which a low-resourced language is mixed with a high resourced language. Low-resourcedness in acoustic data hinders the performance of E2E ASR systems more severely than the conventional ASR systems. To mitigate this problem in the transcription of archives with code-switching Frisian-Dutch speech, we integrate a designated decoding scheme and perform rescoring with neural network-based language models to enable better utilization of the available textual resources. We first incorporate a multi-graph decoding approach which creates parallel search spaces for each monolingual and mixed recognition tasks to maximize the utilization of the textual resources from each language. Further, language model rescoring is performed using a recurrent neural network pre-trained with cross-lingual embedding and further adapted with the limited amount of in-domain CS text. The ASR experiments demonstrate the effectiveness of the described techniques in improving the recognition performance of an E2E CS ASR system in a low-resourced scenario.

This paper presents the design, implementation, and characterization of a hardware platform applicable to a self-powered wireless sensor network (WSN) node. Its primary design objective is to devise a hybrid energy harvesting system to extend the operational lifetime of WSN node after they are deployed in the field environment. Besides the implementation of optimal components (microcontroller, sensor, radio frequency (RF) transceiver, and others) to achieve the lowest power consumption, it is also necessary to consider the sources of energy instead of the frequent recharging or replacement of batteries. Therefore, the platform incorporates a multisource energy harvesting module to collect energy from the surrounding environment, including wind, solar radiation, and thermal energy. The platform also includes an energy storage module through a super-capacitor, RF transceiver module, and the primary microcontroller module. Experimental results showed that the WSN node system with appropriate integration will reserve sufficient energy and meet the long-term power supply requirements of the WSN node without batteries in the field environment. The experimental results and empirical measurements taken over nine days demonstrated that the average daily generating capacity was 7805.09 J, which is far more than the energy consumption of the WSN node (about 2972.88 J).

For multisensor linear time-varying system with non-Gaussian measurement noise, how to design distributed robust estimator to increase the accuracy and robustness to outliers at a relatively low computation and communication cost is a fundamental task. This paper proposes a fast distributed variational Bayesian (VB) filtering algorithm to recursively estimate the state and noise distribution over three conventional sensor networks: 1) incremental-based; 2) diffusion-based; and 3) consensus-based. To be specific, the non-Gaussian measurement noise of each sensor is modeled as Student- t distribution, and the system state and the parameters of the distribution are estimated via VB approach in each iteration step. An interaction scheme is then added to obtain the global optimal parameter by fusing the local optimal parameters over incremental, diffusion, and consensus communication topology. An efficient sensor selection criterion under these topologies based on the Cramér-Rao lower bound is proposed to reduce the communication and computation burden. Compared with the existing centralized VB filtering algorithms, the proposed algorithm in this paper can extensively increase the robustness to node or link failure at a lower computation cost with acceptable estimation performance and communication load. The theoretic results and simulation results are given to show the efficiency of our proposed algorithm.

Dataset plays an essential role in the training and testing of traffic vision algorithms. However, the collection and annotation of images from the real world is time-consuming, labor-intensive, and error-prone. Therefore, more and more researchers have begun to explore the virtual dataset, to overcome the disadvantages of real datasets. In this paper, we propose a systematic method to construct large-scale artificial scenes and collect a new virtual dataset (named “ParallelEye”) for the traffic vision research. The Unity3D rendering software is used to simulate environmental changes in the artificial scenes and generate ground-truth labels automatically, including semantic/instance segmentation, object bounding boxes, and so on. In addition, we utilize ParallelEye in combination with real datasets to conduct experiments. The experimental results show the inclusion of virtual data helps to enhance the per-class accuracy in object detection and semantic segmentation. Meanwhile, it is also illustrated that the virtual data with controllable imaging conditions can be used to design evaluation experiments flexibly.

The haze problem in cities poses a great threat to human health. Although there are many factors that cause fog and haze, the main reason is the increase of the PM2.5concentration. Due to the complexity of the particles motion, it is difficult to use traditional methods obtain information about PM2.5(including its sources, influencing factors, and distribution forecast). This paper presents a cellular automata (CA) model based on a multivariate regression model and several physical models to analyze the generation and diffusion of PM2.5. In Beijing for example, after the researches, the multiple regression confirmed that the major source of PM2.5is vehicle pollution, which accounts for 39.2% of the pollution generated in Beijing. The secondary source is from the residential areas, accounting for 27.5% in the winter. Besides, 32% of the total pollution comes from nearby areas. In addition, it is also confirmed that the weather factors, such as, temperature, wind, pressure and relative humidity, and radiation are all have a great impact on PM2.5. The CA model is demonstrated to be an effective simulation and prediction method for PM2.5since it allows for the estimation of the governance results by simulating the control scheme and predicting the concentration of PM2.5accurately in the next 48 h. In the prediction experiment, 77.5% of prediction error is less than$20~ {\mu }\text{g} / {\text {m}}^{3}$.

In this paper, two multifault diagnosis methods based on improved support vector machine (SVM) are proposed for sensor fault detection and identification respectively. First, online sparse least squares support vector machine (OS-LSSVM) is utilized to detect and predict sensor faults. Then, a method which combines the SVM and error-correcting output codes (ECOC) called ECOC-SVM is proposed to solve the sensor fault feature extraction and online identification problem. We regard nonlinear transformation as the input of classifiers to enhance the separability of initial characteristics. ECOC-SVM is utilized to classify the fault states. Some typical faults are investigated and the experimental results indicate that ECOC-SVM has high identification accuracy and can be implemented in real-time to meet the requirements of online fault identification. This method can also be extended to solve other related problems.

The identification of the nonlinearity and coupling is crucial in nonlinear target tracking problem in collaborative sensor networks. According to the adaptive Kalman filtering (KF) method, the nonlinearity and coupling can be regarded as the model noise covariance, and estimated by minimizing the innovation or residual errors of the states. However, the method requires large time window of data to achieve reliable covariance measurement, making it impractical for nonlinear systems which are rapidly changing. To deal with the problem, a weighted optimization-based distributed KF algorithm (WODKF) is proposed in this paper. The algorithm enlarges the data size of each sensor by the received measurements and state estimates from its connected sensors instead of the time window. A new cost function is set as the weighted sum of the bias and oscillation of the state to estimate the "best" estimate of the model noise covariance. The bias and oscillation of the state of each sensor are estimated by polynomial fitting a time window of state estimates and measurements of the sensor and its neighbors weighted by the measurement noise covariance. The best estimate of the model noise covariance is computed by minimizing the weighted cost function using the exhaustive method. The sensor selection method is in addition to the algorithm to decrease the computation load of the filter and increase the scalability of the sensor network. The existence, suboptimality and stability analysis of the algorithm are given. The local probability data association method is used in the proposed algorithm for the multitarget tracking case. The algorithm is demonstrated in simulations on tracking examples for a random signal, one nonlinear target, and four nonlinear targets. Results show the feasibility and superiority of WODKF against other filtering algorithms for a large class of systems.

In this paper, we present a weighted, hybrid program-dependence model that represents the relevance of highly related, dependent code to assist developer comprehension of the program for multiple software-engineering tasks. Programmers often need to understand the dependencies among program elements, which may exist across multiple modules. Although such dependencies can be gathered from traditional models, such as slices, the scalability of these approaches is often prohibitive for direct, practical use. To address this scalability issue, as well as to assist developer comprehension, we introduce a program model that includes static dependencies as well as information about any number of executions, which inform the weight and relevance of the dependencies. Additionally, classes of executions can be differentiated in such a way as to support multiple software-engineering tasks. We evaluate this weighted, hybrid model for a task that involves exploring the structural context while debugging. The results demonstrate that the new model more effectively reveals relevant failure-correlated code than the static-only model, thus enabling a more scalable exploration or post hoc analysis.

This paper provides techniques for aiding developers' task of familiarizing themselves with the context of a fault. Many fault-localization techniques present the software developer with a subset of the program to inspect in order to aid in the search for faults that cause failures. However, typically, these techniques do not describe how the components of the subset relate to each other in a way that enables the developer to understand how these components interact to cause failures. These techniques also do not describe how the subset relates to the rest of the program in a way that enables the developer to understand the context of the subset. In this paper, we present techniques for providing static and dynamic relations among program elements that can be used as the basis for the exploration of a program when attempting to understand the nature of faults.

Application Programming Interface (API) constraints on objects are rules that API client code must follow in order to get expected results from these objects. Runtime verification, an important approach for detecting API constraint violations, usually suffers from high runtime overhead. This paper focuses on temporal API constraints on multiple interacting objects. Violation detection of such constraints is more challenging than violation detection of single object constraints, and may induce higher runtime overhead. To reduce the runtime overhead, without compromising the effectiveness of verification, we propose a Lazy Verification Approach (LAVA), which enables verification lazily. Verification probes in LAVA are loaded automatically during the program execution as late as possible. And only probes on objects that have been bound by a binding point (a special method invocation that binds involved objects together) are enabled. Based on these optimization strategies, we implemented an efficient and flexible runtime verification framework. We show the effectiveness of our approach by applying it to verify five constraints in the DaCapo [1] benchmark. The empirical results show that our approach can reduce the number of method invocation events sent by probes, which is the main cause of runtime overhead, by 74% to 100% on average, and bring about an optimization ratio of 44.1% to 89.9% on runtime overhead.

Web service technology aims to enable the interoperation of heterogeneous systems and the reuse of distributed functions in an unprecedented scale and has achieved significant success. There are still, however, challenges to realize its full potential. One of these challenges is to ensure the behavior of Web services consistent with their requirements. Monitoring events that are relevant to Web service requirements is, thus, an important technique. This paper introduces an online monitoring approach for Web service requirements. It includes a pattern-based specification of service constraints that correspond to service requirements, and a monitoring model that covers five kinds of system events relevant to client request, service response, application, resource, and management, and a monitoring framework in which different probes and agents collect events and data that are sensitive to requirements. The framework analyzes the collected information against the prespecified constraints, so as to evaluate the behavior and use of Web services. The prototype implementation and experiments with a case study shows that our approach is effective and flexible, and the monitoring cost is affordable.