Search

The work presented in this paper concerns a switching-based control formulation for multi-intersection and multiphase traffic light systems. A macroscopic traffic flow modeling approach is first presented, which is instrumental to the development of a model-based and switching-based optimization method for traffic signal operation, in the framework of adaptive dynamic programming (ADP). The main advantage of the switching-based formulation is its capability to determine both “when”’ to switch and “which” mode to switch on without the need to use the cycle-based average flow approximation typical of state-of-the-art formulations. In addition, the framework can handle different cycle times across intersections without the need for synchronization constraints and, moreover, minimum dwell-time constraints can be directly enforced to comply with minimum green/red times in each phase. The simulation experiments on a multi-intersection and multiphase traffic light systems are presented to show the effectiveness of the method. [ABSTRACT FROM AUTHOR]

Enhancing power electronics (PE) converter reliability is crucial for ensuring a reliable operation of current and future operating Wind Turbines (WTs). Achieving high reliability of variable speed WT PE systems requires careful consideration of their operation, and particularly their thermal cycling. This paper presents a methodology for evaluating and reconsidering operational strategies of WTs with relation to the thermal loading and lifetime consumption of the converter. The methodology is applied to compare control strategies for the WT generator and evaluate their impact on the converter reliability by observation of the thermal cycles and by calculating the resultant lifetime consumption of those stress cycles. The thermal stress on both the Machine Side Converter (MSC) and the Grid Side Converter (GSC) is examined and compared. It is shown that the least reliable of the three evaluated control strategies is the one that tracks the power curve below rated speed most closely. This paper suggests that dynamic transients associated with the WT control largely influence the IGBT module wear-out and their modelling needs to be prioritized for lifetime studies. These dynamic transients are captured by the improved model whose value is confirmed for the comparisons in the case study of the paper. [ABSTRACT FROM AUTHOR]

Accurate retinal fluid segmentation on Optical Coherence Tomography (OCT) images plays an important role in diagnosing and treating various eye diseases. The art deep models have shown promising performance on OCT image segmentation given pixel-wise annotated training data. However, the learned model will achieve poor performance on OCT images that are obtained from different devices (domains) due to the domain shift issue. This problem largely limits the real-world application of OCT image segmentation since the types of devices usually are different in each hospital. In this paper, we study the task of cross-domain OCT fluid segmentation, where we are given a labeled dataset of the source device (domain) and an unlabeled dataset of the target device (domain). The goal is to learn a model that can perform well on the target domain. To solve this problem, in this paper, we propose a novel Structure-guided Cross-Attention Network (SCAN), which leverages the retinal layer structure to facilitate domain alignment. Our SCAN is inspired by the fact that the retinal layer structure is robust to domains and can reflect regions that are important to fluid segmentation. In light of this, we build our SCAN in a multi-task manner by jointly learning the retinal structure prediction and fluid segmentation. To exploit the mutual benefit between layer structure and fluid segmentation, we further introduce a cross-attention module to measure the correlation between the layer-specific feature and the fluid-specific feature encouraging the model to concentrate on highly relative regions during domain alignment. Moreover, an adaptation difficulty map is evaluated based on the retinal structure predictions from different domains, which enforces the model focus on hard regions during structure-aware adversarial learning. Extensive experiments on the three domains of the RETOUCH dataset demonstrate the effectiveness of the proposed method and show that our approach produces state-of-the-art performance on cross-domain OCT fluid segmentation. [ABSTRACT FROM AUTHOR]

Graph-based clustering plays an important role in the clustering area. Recent studies about graph neural networks (GNN) have achieved impressive success on graph-type data. However, in general clustering tasks, the graph structure of data does not exist such that GNN can not be applied to clustering directly and the strategy to construct a graph is crucial for performance. Therefore, how to extend GNN into general clustering tasks is an attractive problem. In this paper, we propose a graph auto-encoder for general data clustering, AdaGAE, which constructs the graph adaptively according to the generative perspective of graphs. The adaptive process is designed to induce the model to exploit the high-level information behind data and utilize the non-euclidean structure sufficiently. Importantly, we find that the simple update of the graph will result in severe degeneration, which can be concluded as better reconstruction means worse update. We provide rigorous analysis theoretically and empirically. Then we further design a novel mechanism to avoid the collapse. Via extending the generative graph models to general type data, a graph auto-encoder with a novel decoder is devised and the weighted graphs can be also applied to GNN. AdaGAE performs well and stably in different scale and type datasets. Besides, it is insensitive to the initialization of parameters and requires no pretraining. [ABSTRACT FROM AUTHOR]

As you may know, the M Barry Carlton Award is bestowed on the authors of the best paper published in the IEEE Transactions of Aerospace and Electronic Systems (T-AES) for a given calendar year. The winners for the 2011 Award were announced in the February issue earlier this year. It is now my privilege to announce that the 2012 M Barry Carlton award winners are Kevin (Jim) Sangston, Fulvio Gini, and Maria (Sabrina) Greco for their paper entitled, "Coherent Radar Target Detection in Heavy-Tailed Compound-Gaussian Clutter." [ABSTRACT FROM AUTHOR]

This paper introduces video domain generalization where most video classification networks degenerate due to the lack of exposure to the target domains of divergent distributions. We observe that the global temporal features are less generalizable, due to the temporal domain shift that videos from other unseen domains may have an unexpected absence or misalignment of the temporal relations. This finding has motivated us to solve video domain generalization by effectively learning the local-relation features of different timescales that are more generalizable, and exploiting them along with the global-relation features to maintain the discriminability. This paper presents the VideoDG framework with two technical contributions. The first is a new deep architecture named the Adversarial Pyramid Network, which improves the generalizability of video features by capturing the local-relation, global-relation, and cross-relation features progressively. On the basis of pyramid features, the second contribution is a new and robust approach of adversarial data augmentation that can bridge different video domains by improving the diversity and quality of augmented data. We construct three video domain generalization benchmarks in which domains are divided according to different datasets, different consequences of actions, or different camera views, respectively. VideoDG consistently outperforms the combinations of previous video classification models and existing domain generalization methods on all benchmarks. [ABSTRACT FROM AUTHOR]

Due to the lack of distinctiveness to the background, robust monocular 6DOF pose tracking of less-distinct objects remains an important open problem. In this paper, we firstly analyze the object distinctiveness in the tracking process. Then, we propose a novel contour-part model based robust monocular pose tracking method for less-distinct objects. This paper uses the traditional contour feature for pose tracking in a novel strategy called contour part model. First, the contour part model is built by segmenting the projected contour rendered from the 3D model into contour segments of a certain length adaptively according to the Shi-Tomasi cornerness scores. Then, the correspondence is detected in the input image for each contour part by gradient orientation based template matching. Finally, pose tracking is achieved by solving the PnP problem. Experiment results on semi-synthetic and real images show that the proposed method performs better than the existing methods when pose tracking less-distinct objects and shows great robustness toward interference. Additionally, we combine the edge feature and the regional feature in a simple strategy. The overall performance of the fused method is boosted according to the experimental results on the RBOT dataset. The fact shows that the fusion of the edge and regional features has great potential for improving tracking performance. [ABSTRACT FROM AUTHOR]

The paper is devoted to the problem of erasure coding in distributed storage. We consider a model of storage that assumes that nodes are organized into equally sized groups, called racks, that within each group the nodes can communicate freely without taxing the system bandwidth, and that the only information transmission that counts is the one between the racks. This assumption implies that the nodes within each of the racks can collaborate before providing information to the failed node. The main emphasis of the paper is on code construction for this storage model. We present an explicit family of maximum distance separable (MDS) array codes that support recovery of a single failed node from any number of helper racks using the minimum possible amount of inter-rack communication(such codes are said to provide optimal repair). The codes are constructed over finite fields of size comparable to the code length. We also derive a bound on the number of symbols accessed at helper nodes for the purposes of repair, and construct a code family that approaches this bound, while still maintaining the optimal repair property. Finally, we present a construction of scalar Reed-Solomon codes that support optimal repair for the rack-oriented storage model. [ABSTRACT FROM AUTHOR]

Three-phase grid-connected inverter modeling depends on the equivalent resistance and inductance between the inverter and the grid. However, these parameters are not fixed during the operation of the inverter and vary with the operating conditions. In this paper, a new globally stable adaptive controller is proposed to estimate these values online and update the controller during its operation. A model reference adaptive system (MRAS) based on two fictitious quantities with no physical significance is utilized, namely $M = v \cdot i^*$ and $N = v^* \times i$ , where $v$ and $i$ are the steady-state values of voltage and current in synchronously rotating frame of reference. Detailed stability analysis of the proposed estimation technique has been presented using a Lyapunov energy function-based approach proving the global stability of the technique. Small-signal modeling and analysis of the proposed scheme has also been presented so as to understand the local dynamics of the system. The guidelines to select the parameter of the proposed MRAS system along with updating the controller are elaborated in this paper. Detailed simulation based on MATLAB/Simulink and PLECS, along with experimental studies through a Texas Instruments TMS320F28377S microcontroller, validates the feasibility of the proposed control architecture. [ABSTRACT FROM AUTHOR]

This paper reconsiders the iterative learning control (ILC) problem for variable trial lengths via compensating output data by using an auxiliary predictive model when the controlled process does not reach the desired trial length. Moreover, this paper aims to propose a general and data-driven ILC method without requiring any explicit mechanistic model information. Specifically, an iterative difference with state transition expression is performed at first over the desired trial length in iteration domain to build an auxiliary predictive model for the iterative input–output dynamics of the linear discrete-time system. Then, an auxiliary predictive compensation-based ILC (APC-ILC) method is presented by defining an expanded output variable in which the predictive output is incorporated to compensate the unavailable output data due to the shorter operation length. The learning gain is iteration-time-varying and is updated using real-time data to adapt to system changes. Furthermore, the proposed learning control law contains additional input information to further improve the control performance. Theoretical analysis and simulations further verify the effectiveness of the proposed APC-ILC. [ABSTRACT FROM AUTHOR]

This paper proposes a model predictive current control (MPCC) method based on current augmenter prediction model (CAPM) appropriate for interior permanent magnet synchronous motor (IPMSM) under high-speed condition. The accuracy of CAPM is improved by considering rotor movement during one control period especially under high-speed condition. Meanwhile, the use of the permanent magnetic flux linkage is eliminated. In order to further improve the robustness of the MPCC method, an adaptive observer based on CAPM is designed to estimate the model error. The proposed adaptive observer also eliminates the use of the permanent magnetic flux linkage, so the whole implementation process of the MPCC method with adaptive observer is independent of the permanent magnetic flux linkage. Moreover, since the adaptive observer with fixed adaptive integral gain cannot achieve satisfactory performance under different operating conditions of the motor, this paper designs a calculation method for the variable adaptive integral gains by considering the current estimation error, the amplitude of current and the angular velocity of the IPMSM. Finally, the proposed method is validated by means of the experimental results on a 20-kW test platform. [ABSTRACT FROM AUTHOR]

Conventionally, the transition probabilities in the interacting multiple model (IMM) are often fixed based on the prior information. However, this conservative setting may result in inaccurate state estimations. To solve this problem, a Bayesian-based online correction function is proposed in this paper, which can adaptively adjust the transition probabilities. To deal with the response lag and the short-term peak estimation error problem during the respond to model jump, a model jumping threshold is defined, so that the current information of the models can be fully utilized by the IMM algorithm and the correction function of the transition probabilities can be further improved. Subsequently, an adaptive transition probability-based parallel IMM algorithm is proposed in this paper. Finally, three maneuvering target tracking simulations are conducted to verify the performance of the proposed algorithm, the results show that the proposed algorithm can improve the response speed of the system model jump and the state estimation accuracy. The effectiveness and feasibility of the algorithm are proven. [ABSTRACT FROM AUTHOR]

Most researchers focus on designing accurate crowd counting models with heavy parameters and computations but ignore the resource burden during the model deployment. A real-world scenario demands an efficient counting model with low-latency and high-performance. Knowledge distillation provides an elegant way to transfer knowledge from a complicated teacher model to a compact student model while maintaining accuracy. However, the student model receives the wrong guidance with the supervision of the teacher model due to the inaccurate information understood by the teacher in some cases. In this paper, we propose a dual-knowledge distillation (DKD) framework, which aims to reduce the side effects of the teacher model and transfer hierarchical knowledge to obtain a more efficient counting model. First, the student model is initialized with global information transferred by the teacher model via adaptive perspectives. Then, the self-knowledge distillation forces the student model to learn the knowledge by itself, based on intermediate feature maps and target map. Specifically, the optimal transport distance is utilized to measure the difference of feature maps between the teacher and the student to perform the distribution alignment of the counting area. Extensive experiments are conducted on four challenging datasets, demonstrating the superiority of DKD. When there are only approximately 6% of the parameters and computations from the original models, the student model achieves a faster and more accurate counting performance as the teacher model even surpasses it. [ABSTRACT FROM AUTHOR]

In this paper, we propose a modified strategy using sparse representation to identify the authorship of online handwritten documents. The note-worthy aspect of this paper lies in the introduction of a-priori information for each dictionary atom, that in a way indicates its degree of importance with respect to the dynamic characteristics of the writer. The methodology behind obtaining this information for the dictionary atoms is based on the computation of entropy values over histograms that are generated from the sparse coefficients during the training phase. The prelearned values are incorporated on the traditional schemes of “max” and “mean” pooling of sparse codes, thereby resulting in a modified writer descriptor. Experiments performed with the proposed sparse classification strategy on the handwritten samples of the IAM and IBM-UB1 online handwriting databases demonstrate promising results. [ABSTRACT FROM AUTHOR]

In fog-aided cellular systems, content delivery latency can be minimized by jointly optimizing edge caching and transmission strategies. In order to account for the cache capacity limitations at the edge nodes (ENs), transmission generally involves both fronthaul transfer from a cloud processor with access to the content library to the ENs and wireless delivery from the ENs to the users. In this paper, the resulting problem is studied from an information-theoretic viewpoint by making the following practically relevant assumptions: 1) the ENs have multiple antennas; 2) only uncoded fractional caching is allowed; 3) the fronthaul links are used to send fractions of contents; and 4) the ENs are constrained to use one-shot linear zero-forcing precoding on the wireless channel. Assuming off-line proactive caching and focusing on a high signal-to-noise ratio (SNR) latency metric, the optimal information-theoretic performance is investigated under both serial and pipelined fronthaul-edge transmission modes. The analysis characterizes the minimum high-SNR latency in terms of normalized delivery time (NDT) for worst case users’ demands. The characterization is exact for a subset of system parameters and is generally optimal within a multiplicative factor of 3/2 for the serial case and 2 for the pipelined case. The results bring insights into the optimal interplay between edge and cloud processing in fog-aided wireless networks as a function of system resources, including the number of antennas at the ENs, the ENs’ cache capacity, and the fronthaul capacity. [ABSTRACT FROM AUTHOR]

The reliable operation of battery management systems depends critically on the accurate estimation of the state of charge (SOC) and characterizing parameters of a battery system. SOC estimation employs models that must be robust against variations in battery cell electrochemical features, aging, and operating conditions. This paper reveals that commonly used SOC estimation schemes are fundamentally flawed in providing the robustness of SOC estimation against model uncertainties. Parameter estimation methodologies and adaptive SOC estimation design are introduced in this paper to enhance SOC estimation accuracy and robustness. By a scrutiny of the impact of parameter variations on SOC estimation accuracy, the SOC–open-circuit-voltage mapping is identified to be the most critical function that must be accurately established. Identification algorithms are introduced, and their convergence properties are established. The integration of the identification algorithms and SOC estimation schemes lead to an adaptive SOC estimation framework that is superior over the existing methods in providing much improved accuracy and robustness. Experimental studies are conducted to validate the algorithms. [ABSTRACT FROM PUBLISHER]

The increasing penetration of distributed renewable energy sources drastically alters the dynamic characteristics of distribution networks (DNs). Therefore, several equivalent models have been recently proposed, to analyze more accurately the complex behavior of modern DNs. However, relatively simple models are still commonly used in practice for dynamic power system studies. In addition, dynamic equivalent models for DNs are sensitive to different operating conditions and there is lack of systematic understanding of their performance. Scope of this paper is to propose a methodology for identifying the applicability range in terms of accuracy and generalization capability of several conventional and newly developed equivalent models for the dynamic analysis of modern DNs. A set of metrics is used for the modelling accuracy assessment and a sensitivity analysis framework is introduced to fully quantify the generalization capabilities of DN equivalent models. Based on the above, guidelines and recommendations for the development of robust equivalent models for DN analysis are proposed. [ABSTRACT FROM AUTHOR]

While recent developments of neural network (NN) models have led to a series of record-breaking achievements in many applications, the lack of sufficiently good datasets remains a problem for some applications. For such a problem, we can however exploit a large number of unstructured text corpora as an external knowledge to complement the training data, and most prevailing neural network solutions employ word embedding methods for such purposes. In this paper, we propose LDA-Reg, a novel knowledge driven regularization framework based on Latent Dirichlet Allocation (LDA) as an alternative to the word embedding methods to adaptively utilize abundant external knowledge and to interpret the NN model. For the joint learning of the parameters, we propose EM-SGD, an effective update method which incorporates Expectation Maximization (EM) and Stochastic Gradient Descent (SGD) to update parameters iteratively. Moreover, we also devise a lazy update and sparse update method for the high-dimensional inputs and sparse inputs respectively. We validate the effectiveness of our regularization framework through an extensive experimental study over real world and standard benchmark datasets. The results show that our proposed framework not only achieves significant improvement over state-of-the-art word embedding methods but also learns interpretable and significant topics for various tasks. [ABSTRACT FROM AUTHOR]

It is widely acknowledged that the success of deep learning is built upon large-scale training data and tremendous computing power. However, the data and computing power are not always available for many real-world applications. In this paper, we address the machine learning problem where it lacks training data and limits computing power. Specifically, we investigate domain adaptation which is able to transfer knowledge from one labeled source domain to an unlabeled target domain, so that we do not need much training data from the target domain. At the same time, we consider the situation that the running environment is confined, e.g., in edge computing the end device has very limited running resources. Technically, we present the Faster Domain Adaptation (FDA) protocol and further report two paradigms of FDA: early stopping and amid skipping. The former accelerates domain adaptation by multiple early exit points. The latter speeds up the adaptation by wisely skip several amid neural network blocks. Extensive experiments on standard benchmarks verify that our method is able to achieve the comparable and even better accuracy but employ much less computing resources. To the best of our knowledge, there are very few works which investigated accelerating knowledge adaptation in the community. This work is expected to inspire the topic for more discussion. [ABSTRACT FROM AUTHOR]

In this paper, adaptive quantized state estimation fusion is deeply studied. To approach the model mismatching problem induced by random quantization, some quantized Kalman filters have been presented in the previous work, such as the quantized Kalman filter with strong tracking filtering (QKF-STF), the variational Bayesian adaptive quantized Kalman filter (VB-AQKF), and a centralized fusion frame-based complex quantized filter called variational Bayesian adaptive QKF-STF (VB-AQKF-STF). Based on the previous work for the single sensor system, a distributed complex quantized filter is designed in this paper. A novel quantized Kalman filter based on multiple-method fusion scheme (QKF-MMF) is proposed. Similar to the VB-AQKF-STF, the QKF-MMF can also realize joint estimation on the state and the quantization error covariance under the distributed fusion frame. Furthermore, it extends the single sensor results to multisensor tracking systems by using centralized and distributed fusion frames. Two multisensor quantized fusion estimators are proposed for a parallel structure with main-secondary processors in the fusion center. The weighted fusion and embedded integration ways are deeply applied to design the multisensor quantized fusion methods. The proposed work can perfect the quantized estimation algorithms and provide different choices for practical engineering applications. [ABSTRACT FROM AUTHOR]

During the development or deployment of discrete event simulation systems, many sudden changes in requirements emerge. To embrace the changes rapidly and reduce development costs, attaching reusable, and black-boxed components to existing systems has been regarded as one of the most effective approaches because modifications at the code level are generally costly and risky. However, since this approach requires components that are an exact fit, it may not be easy to avoid the modifications at the source code level actually. Moreover, the required components may not exist. Hence, this paper applies black-box extensibility to simulation systems in order to avoid the modifications. This paper proposes adaptive discrete event simulation systems using event control models for extending and modifying semantics through event-based simulation interface. With the proposed work, the simulation events are modulated, deleted, and generated by the event-oriented control functions in the event control model to embrace the changes of requirements. It can substitute for the modifications at the code level and extend the existing behavioral semantics. As a result, the proposed work provides a new alternative step in the development process with reusable components to avoid modifications at the source code level. The new step will lead to rapid adaptations of existing simulation systems. To support the effectiveness of this approach, this paper will describe applicable examples based on our empirical studies. [ABSTRACT FROM AUTHOR]

In this paper, we modify the graph model for conflict resolution (GMCR) to model interactive unawareness of decision makers (DMs) about the options available to them in the conflict. More specifically, we consider a GMCR with two DMs, where a DM, in some given state, can be unconscious about some of his options, or about the options of his opponent, and therefore, may have only a partial knowledge of the state space. By interactive unawareness, it is meant that a DM can reason about the awareness level of the other DM and about the awareness level of the other DM regarding his or her own awareness level and so on. We generalize standard solution concepts for this model and illustrate its usefulness by means of a hypothetical war conflict. [ABSTRACT FROM AUTHOR]

In this paper, we develop robust adaptive nonsingular terminal sliding mode (NTSM) control methodologies to solve the position and the velocity tracking control problem of the automatic train operation (ATO) system subject to unknown parameters, model uncertainty, and external disturbances. A novel nonlinear nonsingular terminal sliding manifold is proposed by considering that its parameter is unknown, which need to be estimated via a proposed non-negative adaptive law. And a corresponding novel robust adaptive NTSM control strategy, which enables the position tracking error and the velocity tracking error of the ATO system to converge to zero, and eliminates the singularity caused by terminal sliding mode controller, is proposed. Furthermore, unknown parameters of the sliding manifold and the ATO system can be estimated online by the proposed methodology. Simulation results show the effectiveness of the proposed methodologies in this paper. [ABSTRACT FROM AUTHOR]

This paper proposes a bioinspired reference model based fuzzy adaptive tracking control for active suspension systems. A general bioinspired nonlinear structure, which can present ideal nonlinear quasi-zero stiffness for vibration isolation, is adopted as tracking reference model. Fuzzy logic systems are used to approximate unknown nonlinear terms in nonlinear suspension systems. Particularly, a nonlinear damping is designed to improve damping characteristics of the bioinspired reference model. With beneficial nonlinear stiffness and improved nonlinear damping of the bioinspired reference model, the proposed fuzzy adaptive controller can effectively suppress vibration of suspension systems with less actuator force and much improved ride comfort, thus energy saving performance can be achieved. Finally, a quarter-vehicle active suspension system with considering payload uncertainties, general disturbance, and actuator saturation is provided for evaluating the validity and superiority of the bioinspired nonlinear dynamics based fuzzy adaptive control approach proposed in this paper. [ABSTRACT FROM AUTHOR]

In this paper, we quantify the impact of uncertainty in predicted mobility information on the sensing cost of the mobile sensing problem (i.e. sensing of target areas) in vehicular sensor networks. We focus on vehicular sensor networks in which the communication channel is the bottleneck of the system, and sensing the environment is meant to be provided with a minimum load on the communication channel. First, we formulate an opportunistic scheduler, which does not utilize predicted mobility information in the sensing procedure. Then, we formulate a strategic scheduler, which utilizes predicted mobility information in the sensing procedure. After that, we propose two types of noise models that capture the uncertainty in knowing mobility information. We quantify the impact of uncertainty in knowing mobility information on the sensing cost of the strategic scheduler using an independent coverage model. We find that the strategic scheduler outperforms the noise-free opportunistic scheduler even when the strategic scheduler utilizes noisy mobility information up to a certain threshold of noise. This threshold is determined in the paper, and is referred to as the breaking point of the strategic scheduler. Simulation results of the breaking point closely matches our analysis, which demonstrates the correctness of the used analysis methodology. Finally, we use an estimation model for the mobility information between two points on the road, and we propose an algorithm to keep the scheduling scheme below the breaking point. [ABSTRACT FROM AUTHOR]

In this paper, we are interested in the compression of image sets or video with considerable changes in illumination. We develop a framework to decompose frames into illumination fields and texture in order to achieve sparser representations of frames which is beneficial for compression. Illumination variations or contrast ratio factors among frames are described by a full resolution multiplicative field. First, we propose a Lifting-based Illumination Adaptive Transform (LIAT) framework which incorporates illumination compensation to temporal wavelet transforms. We estimate a full resolution illumination field, taking heed of its spatial sparsity by a rate-distortion (R-D) driven framework. An affine mesh model is also developed as a point of comparison. We find the operational coding cost of the subband frames by modeling a typical t + 2D wavelet video coding system. While our general findings on R-D optimization are applicable to a range of coding frameworks, in this paper, we report results based on employing JPEG 2000 coding tools. The experimental results highlight the benefits of the proposed R-D driven illumination estimation and compensation in comparison with alternative scalable coding methods and non-scalable coding schemes of AVC and HEVC employing weighted prediction. [ABSTRACT FROM AUTHOR]

This paper presents a robust nonlinear decentralized control scheme for islanded dc microgrids, where the main control objectives are to achieve the desired voltage at the common dc bus and to maintain the power balance. The proposed control scheme uses the partial feedback linearization scheme to simplify the dynamical models of different components in dc microgrids. In this paper, the dc microgrid includes a solar photovoltaic unit, a fuel cell system, and a battery energy storage system along with dc loads. The robustness of the proposed controller against parametric uncertainties is ensured by the parameter, appearing in the control inputs as unknown, which is then estimated through adaptation laws. The inherent noise decoupling capability of the feedback linearization scheme is used to provide robustness against external disturbances in dc microgrids. The performance of the proposed controller is evaluated on a dc microgrid through simulation and experimental studies in order to demonstrate the effectiveness and robustness under different operating conditions, while considering the effects of parametric uncertainties and external disturbances. Simulation results are also compared with an existing proportional-integral controller. [ABSTRACT FROM AUTHOR]

In this paper, a simplified model of a 2 × 25 kV bi-level traction power system for feeding high-speed trains is presented. A modified nodal analysis (MNA) based algorithm for solving the system with and adaptive damping factor is also explained. The main novelties of this paper are the inclusion of train protections (over-current and over-voltage) in the model and the development of the MNA solving procedure that increases the stability and robustness of the iterative solving process. Another important feature of the model is that it can be easily adapted to a single voltage feeding systems by deactivation of specific parameters. The accuracy and performance of the proposed simulator is compared and verified relative to the derivative based solvers. [ABSTRACT FROM AUTHOR]

This paper proposes a novel approach for power system dynamic simulation based on the differential transformation (DT). The DT is introduced to study power systems as high-dimensional nonlinear dynamical systems for the first time, and is able to avoid computations of high-order derivatives with nonlinear differential equations by its transform rules. This paper, first, proposes and proves several new transform rules for generic nonlinear functions that often appear in power system models, and then uses these rules to transform representative power system models such as the synchronous machine model with trigonometric functions and the exciter model with exponential and square root functions. This paper also designs a DT-based simulation scheme that allows significantly prolonged time steps to reduce simulation time compared to a traditional numerical approach. The numerical stability, accuracy, and time performance of the proposed new DT-based simulation approach are compared with widely used numerical methods on the IEEE 39-bus system and Polish 2383-bus system. [ABSTRACT FROM AUTHOR]

In underwater acoustic (UWA) adaptive communication system, due to time-varying channel, the transmitter often has outdated channel state information (CSI), which results in low efficiency. UWA channels are much more difficult to estimate and predict than terrestrial wireless channels, given the more severe multipath environments with varying propagation speeds in different locations, non-linear propagation paths, several-order higher propagation latency, mobile transceiver and obstacles in the sea, etc. To handle the complexity, this paper proposes an efficient online CSI prediction model for UWA CSI prediction considering the complicated correlationship of UWA channels in both the time and frequency domains. This paper designs a learning model called CsiPreNet, which is an integration of a one-dimensional convolutional neural network (CNN) and a long short term memory (LSTM) network. The performance is compared with the widely used recursive least squares (RLS) predictor, the approximate linear dependency recursive kernel least-squares (ALD-KRLS), and two common conventional deep neural networks (DNN) predictors, i.e., back propagation neural network (BPNN) and LSTM network using the measured data recorded in the South China Sea. To validate the efficacy of prediction, we investigate the prediction of CSI in simulated downlink UWA orthogonal frequency division multiple access (OFDMA) systems. Specifically, the measured UWA channel is used in the OFDMA system. A joint subcarrier-bit-power adaptive allocation scheme is used for resource allocation. To further improve the performance, we develop an offline-online prediction scheme, enabling the prediction results to be more stable. Simulation and experimental results show that the CsiPreNet has superior performance than the existing solutions, thanks to its capability in capturing both the temporal and frequency correlation of the UWA CSIs. [ABSTRACT FROM AUTHOR]

This paper proposes a sampled data-based modeling approach for complex network systems. A compression method, known as characteristic modeling, is used to construct microscopic models from macroscopic observations. Based on this model, a novel control method is also developed to achieve network synchronization. The proposed approach can reduce both the complexity of microscopic dynamics and overall networks. Its application to pinning control design validates the effectiveness of this approach. [ABSTRACT FROM AUTHOR]

Effective transient power grid simulators are needed during design processes because a designed power grid needs to be numerously analyzed. This paper builds an efficient and reliable incremental power grid transient simulator, which we name it InTraSim, by integrating macro modeling techniques, sparse recovery mechanisms, an innovated pseudo-node value estimation method, and an initiated adaptive error control scheme. InTraSim is able to deal with not only element-value alterations of designs but also topology modifications of designs. [ABSTRACT FROM PUBLISHER]

This paper critically analyzes several incoherent model-based decomposition methods for assessing the effect of deorientation in characterization of various land covers. It has been found that even after performing decomposition, ambiguity still occurs in scattering response from various land covers, such as urban and vegetation. Researchers introduced the concept of deorientation to remove this ambiguity. Therefore, in this paper, a critical analysis has been carried out using seven different three- and four-component decomposition methods with and without deorientation and two Eigen decomposition-based methods to investigate the scattering response on various land covers, such as urban, vegetation, bare soil, and water. The comprehensive evaluation of decomposition and deorientation effect has been performed by both visual and quantitative analyses. Two types of quantitative analysis have been performed; first, by observing percentage of scattering power and second, by analyzing the variation in the number of pixels in different land covers for each scattering contribution. The analysis shows that deorientation increases not only the power but also the number of pixels for surface and double bounce scattering. The number of pixels representing volume scattering remain almost the same for all the methods with or without deorientation, whereas volume scattering power reduces after deorientation. Eigen decomposition-based methods are observed to solve the problem of overestimation of volume scattering power. [ABSTRACT FROM PUBLISHER]

In recent years, an emerging stream of research suggests that firms should use technological innovation and business model innovation harmoniously to maximize firm performance. In order to enrich the scholarly conversation on the fit between technological innovation and business model innovation, drawing insights from the dynamic capabilities perspective, this paper examines the fit between value proposition innovation and technological innovation (exploitative versus explorative) for the performance of startups in the digital environment. Based on on-site survey data of 285 digital startups in one of the world's largest digital economies, we find that explorative innovation strengthens the positive impact of value proposition innovation on the performance of startups, whereas exploitative innovation weakens this positive effect. Furthermore, these moderating effects are magnified in a highly uncertain demand environment. These findings inject fresh insights into existing scholarly conversation on the fit between technological innovation and business model innovation by offering a dynamic capabilities perspective and by extending it to an increasingly digital business environment. [ABSTRACT FROM AUTHOR]

With the boom of deep neural networks, blind image quality assessment (BIQA) has achieved great processes. However, the current BIQA metrics are limited when evaluating low-quality images as compared to medium-quality and high-quality images, which restricts their applications in real world problems. In this paper, we first identify that two challenges caused by distribution shift and long-tailed distribution lead to the compromised performance on low-quality images. Then, we propose an intermediary enhancement-based bilateral network with iterative training strategy for solving these two challenges. Drawing on the experience of transitive transfer learning, the proposed metric adaptively introduces enhanced intermediary images to transfer more information to low-quality images for mitigating the distribution shift. Our metric also adopts an iterative training strategy to deal with the long-tailed distribution. This strategy decouples feature extraction and score regression for better representation learning and regressor training. It not only transfers the knowledge learned from the earlier stage to the latter stage, but also makes the model pay more attention to long-tailed low-quality images. We conduct extensive experiments on five authentically distorted image quality datasets. The results show that our metric significantly improves the evaluating performance on low-quality images and delivers state-of-the-art intra-dataset results. During generalization tests, our metric also achieves the best cross-dataset performance. [ABSTRACT FROM AUTHOR]

Video Anomaly detection in videos refers to the identification of events that do not conform to expected behavior. However, almost all existing methods cast this problem as the minimization of reconstruction errors of training data including only normal events, which may lead to self-reconstruction and cannot guarantee a larger reconstruction error for an abnormal event. In this paper, we propose to formulate the video anomaly detection problem within a regime of video prediction. We advocate that not all video prediction networks are suitable for video anomaly detection. Then, we introduce two principles for the design of a video prediction network for video anomaly detection. Based on them, we elaborately design a video prediction network with appearance and motion constraints for video anomaly detection. Further, to promote the generalization of the prediction-based video anomaly detection for novel scenes, we carefully investigate the usage of a meta learning within our framework, where our model can be fast adapted to a new testing scene with only a few starting frames. Extensive experiments on both a toy dataset and three real datasets validate the effectiveness of our method in terms of robustness to the uncertainty in normal events and the sensitivity to abnormal events. [ABSTRACT FROM AUTHOR]

Pretrained deep models hold their learnt knowledge in the form of model parameters. These parameters act as “memory” for the trained models and help them generalize well on unseen data. However, in absence of training data, the utility of a trained model is merely limited to either inference or better initialization towards a target task. In this paper, we go further and extract synthetic data by leveraging the learnt model parameters. We dub them Data Impressions, which act as proxy to the training data and can be used to realize a variety of tasks. These are useful in scenarios where only the pretrained models are available and the training data is not shared (e.g., due to privacy or sensitivity concerns). We show the applicability of data impressions in solving several computer vision tasks such as unsupervised domain adaptation, continual learning as well as knowledge distillation. We also study the adversarial robustness of lightweight models trained via knowledge distillation using these data impressions. Further, we demonstrate the efficacy of data impressions in generating data-free Universal Adversarial Perturbations (UAPs) with better fooling rates. Extensive experiments performed on benchmark datasets demonstrate competitive performance achieved using data impressions in absence of original training data. [ABSTRACT FROM AUTHOR]

Although deep face recognition has achieved impressive progress in recent years, controversy has arisen regarding discrimination based on skin tone, questioning their deployment into real-world scenarios. In this paper, we aim to systematically and scientifically study this bias from both data and algorithm aspects. First, using the dermatologist approved Fitzpatrick Skin Type classification system and Individual Typology Angle, we contribute a benchmark called Identity Shades (IDS) database, which effectively quantifies the degree of the bias with respect to skin tone in existing face recognition algorithms and commercial APIs. Further, we provide two skin-tone aware training datasets, called BUPT-Globalface dataset and BUPT-Balancedface dataset, to remove bias in training data. Finally, to mitigate the algorithmic bias, we propose a novel meta-learning algorithm, called Meta Balanced Network (MBN), which learns adaptive margins in large margin loss such that the model optimized by this loss can perform fairly across people with different skin tones. To determine the margins, our method optimizes a meta skewness loss on a clean and unbiased meta set and utilizes backward-on-backward automatic differentiation to perform a second order gradient descent step on the current margins. Extensive experiments show that MBN successfully mitigates bias and learns more balanced performance for people with different skin tones in face recognition. The proposed datasets are available at http://www.whdeng.cn/RFW/index.html. [ABSTRACT FROM AUTHOR]

Unsupervised domain adaptation (UDA) aims to transfer knowledge from a related but different well-labeled source domain to a new unlabeled target domain. Most existing UDA methods require access to the source data, and thus are not applicable when the data are confidential and not shareable due to privacy concerns. This paper aims to tackle a realistic setting with only a classification model available trained over, instead of accessing to, the source data. To effectively utilize the source model for adaptation, we propose a novel approach called Source HypOthesis Transfer (SHOT), which learns the feature extraction module for the target domain by fitting the target data features to the frozen source classification module (representing classification hypothesis). Specifically, SHOT exploits both information maximization and self-supervised learning for the feature extraction module learning to ensure the target features are implicitly aligned with the features of unseen source data via the same hypothesis. Furthermore, we propose a new labeling transfer strategy, which separates the target data into two splits based on the confidence of predictions (labeling information), and then employ semi-supervised learning to improve the accuracy of less-confident predictions in the target domain. We denote labeling transfer as SHOT++ if the predictions are obtained by SHOT. Extensive experiments on both digit classification and object recognition tasks show that SHOT and SHOT++ achieve results surpassing or comparable to the state-of-the-arts, demonstrating the effectiveness of our approaches for various visual domain adaptation problems. Code will be available at https://github.com/tim-learn/SHOT-plus. [ABSTRACT FROM AUTHOR]

Domain adaptation aims at extracting knowledge from auxiliary source domains to assist the learning task in a target domain. In classification problems, since the distributions of the source and target domains are different, directly using source data to build a classifier for the target domain may hamper the classification performance on the target data. Fortunately, in many tasks, there can be some features that are transferable, i.e., the source and target domains share similar properties. On the other hand, it is common that the source data contain noisy features which may degrade the learning performance in the target domain. This issue, however, is barely studied in existing works. In this paper, we propose to find a feature subset that is transferable across the source and target domains. As a result, the domain discrepancy measured on the selected features can be reduced. Moreover, we seek to find the most discriminative features for classification. To achieve the above goals, we formulate a new sparse learning model that is able to jointly reduce the domain discrepancy and select informative features for classification. We develop two optimization algorithms to address the derived learning problem. Extensive experiments on real-world data sets demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Short text clustering has far-reaching effects on semantic analysis, showing its importance for multiple applications such as corpus summarization and information retrieval. However, it inevitably encounters the severe sparsity of short text representations, making the previous clustering approaches still far from satisfactory. In this paper, we present a novel attentive representation learning model for shot text clustering, wherein cluster-level attention is proposed to capture the correlations between text representations and cluster representations. Relying on this, the representation learning and clustering for short texts are seamlessly integrated into a unified model. To further ensure robust model training for short texts, we apply adversarial training to the unsupervised clustering setting, by injecting perturbations into the cluster representations. The model parameters and perturbations are optimized alternately through a minimax game. Extensive experiments on four real-world short text datasets demonstrate the superiority of the proposed model over several strong competitors, verifying that robust adversarial training yields substantial performance gains. [ABSTRACT FROM AUTHOR]

Severe contrast degradation in an image result in a substantial loss of crucial visible information. A novel unsupervised image enhancement algorithm is proposed in this paper to unravel hidden visual details in perceptually invisible images. In this paper, new algorithms are designed to solve the range localization problem. The proposed algorithm intends to expand the image range to maximize its intensity level utilization. Outliers present in an image intensity range are truncated based on their power distribution. Based on the type of input image, maxima-minima, or frequency-based expansion of truncated levels, is exercised to avoid over-enhancement. Two novel parameters, namely the transmission map and zero-frequency spread, are coined. The visually enhanced result successfully exhibits the camouflaged information and avoids over-enhancement. Consumer gadgets, such as cameras, have difficulty taking pictures in very dark and cloudy settings. The suggested algorithm can improve low-contrast photos that are both very dark and severely attenuated. The unsupervised nature of the proposed algorithm makes it suitable for application in the field of consumer electronics like mobile phones and digital cameras. A consumer-focused, platform-independent app is created and made available to the general public for additional study. [ABSTRACT FROM AUTHOR]

As power systems are undergoing a significant transformation with more uncertainties, less inertia and closer to operation limits, there is increasing risk of large outages. Thus, there is an imperative need to enhance grid emergency control to maintain system reliability and security. Towards this end, great progress has been made in developing deep reinforcement learning (DRL) based grid control solutions in recent years. However, existing DRL-based solutions have two main limitations: 1) they cannot handle well with a wide range of grid operation conditions, system parameters, and contingencies; 2) they generally lack the ability to fast adapt to new grid operation conditions, system parameters, and contingencies, limiting their applicability for real-world applications. In this paper, we mitigate these limitations by developing a novel deep meta-reinforcement learning (DMRL) algorithm. The DMRL combines the meta strategy optimization together with DRL, and trains policies modulated by a latent space that can quickly adapt to new scenarios. We test the developed DMRL algorithm on the IEEE 300-bus system. We demonstrate fast adaptation of the meta-trained DRL polices with latent variables to new operating conditions and scenarios using the proposed method, which achieves superior performance compared to the state-of-the-art DRL and model predictive control (MPC) methods. [ABSTRACT FROM AUTHOR]

The Czochralski (CZ) process is the core technology for producing semiconductor silicon monocrystalline (SMC), and it is a complex batch process. However, the crystal growth rate and crystal diameter, which are key quality indicators, are difficult to detect directly online, and the offline calculation process lags seriously, which easily causes blind crystal quality control. Therefore, this paper proposes a data-driven and mechanism-based hybrid model for semiconductor SMC quality variables prediction in the CZ process. Firstly, a data-driven model JITL-SAE-ELM based on just-in-time learning (JITL) fine-tuning strategy is proposed. This model is used to solve the nonlinear and time-varying relationship of the energy transfer process in the CZ process that cannot be accurately described by traditional mechanism models. Here, the SAE-ELM model integrates a stacked autoencoder (SAE) and an extreme learning machine (ELM), which are used to deeply capture the nonlinear and time-varying features of process data. Secondly, according to the hydrodynamics and geometric behavior, a crystal pulling dynamic mechanism model based on the crystal growth mechanism is constructed, which avoids the complicated heat transfer link. Further, considering the unmodeled dynamics caused by model parameter uncertainty during the combination of the energy transfer model and crystal pulling dynamic mechanism model, a crystal diameter compensation model SAE-ELM was developed to improve the prediction accuracy of the CZ process hybrid model. Finally, an industrial data experiment based on a CZ monocrystal furnace illustrates the proposed method. [ABSTRACT FROM AUTHOR]

Few-shot learning is an extremely challenging task in computer vision that has attracted increased research attention in recent years. However, most recent methods do not fully use the task’s information, and few of the seen samples result in large intraclass differences among the same classes. In this paper, we propose a novel task encoding with distribution calibration (TEDC) model for few-shot learning, which uses the relationships among the feature distributions to reduce intraclass differences. In the TEDC model, an integrated feature extraction module (IFEM) is proposed, which extracts the multiangle visual features of an image and fuses them to obtain more representative features. To effectively utilize the task information, a novel task encoding module (TEM) is proposed, which obtains the task features by fusing all the seen samples’ information and uses them to adjust all the samples’ features for more generalizable task-specific representations. We also propose a distribution calibration module (DCM) to reduce the bias between the distribution of the support features and the query features in the same class. Extensive experiments show that our proposed TEDC model achieves an excellent performance and outperforms the state-of-the-art methods on three widely used few-shot classification benchmarks, specifically miniImageNet, tieredImageNet and CUB-200-2011. [ABSTRACT FROM AUTHOR]

Although hand pose estimation has achieved a great success in recent years, there are still challenges with RGB-based estimation tasks, the most significant of which is the absence of labeled training data. At present, the synthetic dataset has plenty of images with accurate annotation, but the difference from real-world datasets affects generalization. Therefore, a transfer learning strategy, which tries to transfer knowledge from a labeled source domain to an unlabeled target domain, is a frequent solution. Existing methods such as mean-teacher, Cyclegan, and MCD will train models with the help of some easily accessible domains such as synthetic data. However, these methods are not guaranteed to operate well in real-world settings due to the domain shift. In this paper, we design a new unsupervised domain adaptation method named Multi-branch Adversarial Regressors (MarsDA) in hand pose estimation, where it could be better for feature migration. Specifically, we first generate pseudo-labels for unlabeled target domain data. Then, the new adversarial training loss between multiple regression branches we designed for hand pose estimation is introduced to narrow the domain gap. In this way, our model can reduce the noise of pseudo labels caused by the domain gap and improve the accuracy of pseudo labels. We evaluate our method on two publicly available real-world datasets, H3D and STB. Experimental results show that our method outperforms existing methods by a large margin. [ABSTRACT FROM AUTHOR]

Histopathological tissue classification is a simpler way to achieve semantic segmentation for the whole slide images, which can alleviate the requirement of pixel-level dense annotations. Existing works mostly leverage the popular CNN classification backbones in computer vision to achieve histopathological tissue classification. In this paper, we propose a super lightweight plug-and-play module, named Pyramidal Deep-Broad Learning (PDBL), for any well-trained classification backbone to improve the classification performance without a re-training burden. For each patch, we construct a multi-resolution image pyramid to obtain the pyramidal contextual information. For each level in the pyramid, we extract the multi-scale deep-broad features by our proposed Deep-Broad block (DB-block). We equip PDBL in three popular classification backbones, ShuffLeNetV2, EfficientNetb0, and ResNet50 to evaluate the effectiveness and efficiency of our proposed module on two datasets (Kather Multiclass Dataset and the LC25000 Dataset). Experimental results demonstrate the proposed PDBL can steadily improve the tissue-level classification performance for any CNN backbones, especially for the lightweight models when given a small among of training samples (less than 10%). It greatly saves the computational resources and annotation efforts. The source code is available at: https://github.com/linjiatai/PDBL. [ABSTRACT FROM AUTHOR]