Search

The performance of a convolutional neural network (CNN) heavily depends on its hyperparameters. However, finding a suitable hyperparameters configuration is difficult, challenging, and computationally expensive due to three issues, which are 1) the mixed-variable problem of different types of hyperparameters; 2) the large-scale search space of finding optimal hyperparameters; and 3) the expensive computational cost for evaluating candidate hyperparameters configuration. Therefore, this article focuses on these three issues and proposes a novel estimation of distribution algorithm (EDA) for efficient hyperparameters optimization, with three major contributions in the algorithm design. First, a hybrid-model EDA is proposed to efficiently deal with the mixed-variable difficulty. The proposed algorithm uses a mixed-variable encoding scheme to encode the mixed-variable hyperparameters and adopts an adaptive hybrid-model learning (AHL) strategy to efficiently optimize the mixed-variables. Second, an orthogonal initialization (OI) strategy is proposed to efficiently deal with the challenge of large-scale search space. Third, a surrogate-assisted multi-level evaluation (SME) method is proposed to reduce the expensive computational cost. Based on the above, the proposed algorithm is named surrogate-assisted hybrid-model EDA (SHEDA). For experimental studies, the proposed SHEDA is verified on widely used classification benchmark problems, and is compared with various state-of-the-art methods. Moreover, a case study on aortic dissection (AD) diagnosis is carried out to evaluate its performance. Experimental results show that the proposed SHEDA is very effective and efficient for hyperparameters optimization, which can find a satisfactory hyperparameters configuration for the CIFAR10, CIFAR100, and AD diagnosis with only 0.58, 0.97, and 1.18 GPU days, respectively.

Background: We aimed to develop census-linked longitudinal mortality data for Japan and assess their validity as a new resource for estimating socioeconomic inequalities in health. Methods: Using deterministic linkage, we identified, from national censuses for 2000 and 2010 and national death records, persons and deceased persons who had unique personal identifiers (generated using sex, birth year/month, address, and marital status). For the period 2010–2015, 1,537,337 Japanese men and women aged 30–79 years (1.9% in national census) were extracted to represent the sample population. This population was weighted to adjust for confounding factors. We estimated age-standardized mortality rates (ASMRs) by education level and occupational class. The slope index of inequality (SII) and relative index inequality (RII) by educational level were calculated as inequality measures. Results: The reweighted sample population’s mortality rates were somewhat higher than those of the complete registry, especially in younger age-groups and for external causes. All-cause ASMRs (per 100,000 person-years) for individuals aged 40–79 years with high, middle, and low education levels were 1,078 (95% confidence interval [CI], 1,051–1,105), 1,299 (95% CI, 1,279–1,320), and 1,670 (95% CI, 1,634–1,707) for men, and 561 (95% CI, 536–587), 601 (95% CI, 589–613), and 777 (95% CI, 745–808) for women, respectively, during 2010–2015. SII and RII by educational level increased among both sexes between 2000–2005 and 2010–2015, which indicates that mortality inequalities increased. Conclusion: The developed census-linked longitudinal mortality data provide new estimates of socioeconomic inequalities in Japan that can be triangulated with estimates obtained with other methods.

Issuing a disaster certificate, which is used to decide the contents of a victim’s support, requires accuracy and rapidity. However, in Japan at large, issuing of damage certificates has taken a long time in past earthquake disasters. Hence, the government needs a more efficient mechanism for issuing damage certificates. This study developed an estimation system of roof-damaged buildings to obtain an overview of earthquake damage based on aero-photo images using deep learning. To provide speedy estimation, this system utilized the trimming algorithm, which automatically generates roof image data using the location information of building polygons on GIS (Geographic Information System). Consequently, the proposed system can estimate, if a house is covered with a blue sheet with 97.57 % accuracy and also detect whether a house is damaged, with 93.51 % accuracy. It would therefore be worth considering the development of an image recognition model and a method of collecting aero-photo data to operate this system during a real earthquake.

We propose a novel two-stage training strategy with ambiguity boosting for the self-supervised learning of single view depths from stereo images. Our proposed two-stage learning strategy firstly aims to obtain a coarse depth prior by training an auto-encoder network for a stereoscopic view synthesis task. This prior knowledge is then boosted and used to self-supervise the model in the second stage of training in our novel ambiguity boosting loss. Our ambiguity boosting loss is a confidence-guided type of data augmentation loss that improves the accuracy and consistency of generated depth maps under several transformations of the single-image input. To show the benefits of the proposed two-stage training strategy with boosting, our two previous depth estimation (DE) networks, one with t-shaped adaptive kernels and the other with exponential disparity volumes, are extended with our new learning strategy, referred to as DBoosterNet-t and DBoosterNet-e, respectively. Our self-supervised DBoosterNets are competitive, and in some cases even better, compared to the most recent supervised SOTA methods, and are remarkably superior to the previous self-supervised methods for monocular DE on the challenging KITTI dataset. We present intensive experimental results, showing the efficacy of our method for the self-supervised monocular DE task.

Objectives Mild Cognitive Impairment (MCI) is the prodromal stage of Alzheimer's disease (AD), which is a progressive and fatal neurodegenerative disorder. Detection of MCI condition can enable early diagnosis resulting in timely intervention to delay the disease progression. Onset of MCI causes tissue alterations in Corpus Callosum (CC) of the brain. Texture analysis of brain Magnetic Resonance (MR) images aids in characterising these imperceptible changes. In this study, Kernel Density Estimation (KDE) technique is used to analyse the textural variations in CC to detect MCI condition. Materials and method The pre-processed brain MR images are obtained from a public access database. Reaction Diffusion level set is employed to segment CC from sagittal slices of the images. Kernel density estimation method is applied to study the local intensity variations within the segmented CC. Statistical features quantifying these variations are extracted from the KDE values. These features are used to differentiate MCI condition using linear classifiers based on discriminant analysis and support vector machine. The results are compared with conventional Grey Level Co-occurrence Matrix (GLCM) features for validation. Results The KDE-based texture features extracted from CC show significant variation between normal and MCI classes. Results demonstrate that this approach can differentiate MCI condition with high accuracy and specificity of 81.3% and 82.7%, respectively. The KDE-based features perform better when compared with GLCM features for distinguishing MCI. Conclusions The KDE-based texture features are able to capture the subtle changes occurring in CC at the MCI stage. This technique achieves comparable performance to other state-of-the-art methods with reduced number of features. Efficiency of the KDE-based texture analysis confirms that the proposed computer assisted technique can be used for mass screening of MCI, which can aid in handling the disease severity.

Machine learning research, particularly in genomics, is often based on wide shaped datasets, i.e. datasets having a large number of features, but a small number of samples. Such configurations raise the possibility of chance influence (the increase of measured accuracy due to chance correlations) on the learning process and the evaluation results. Prior research underlined the problem of generalization of models obtained based on such data. In this paper, we investigate the influence of chance on prediction and show its significant effects on wide shaped datasets. First, we empirically demonstrate how significant the influence of chance in such datasets is by showing that prediction models trained on thousands of randomly generated datasets can achieve high accuracy. This is the case even when using cross-validation. We then provide a formal analysis of chance influence and design formal chance influence estimators based on the dataset parameters, namely its sample size, the number of features, the number of classes and the class distribution. Finally, we provide an in-depth discussion of the formal analysis including applications of the findings and recommendations on chance influence mitigation.

A secure distributed estimation method against data integrity attacks is presented for general Internet-of-Things (IoT) systems. By exploiting the spatial sparsity of the attacks, a robust optimal estimation objective to protect the entire IoT system is established. Then, a false data processor is developed to suppress the bad effects of the attacks. The convergence of the presented method is analyzed. It shows that, under accessible conditions, the estimation error will converge to be uniformly bounded for all cases, i.e., all communication links and transmitted data are arbitrarily compromised. Furthermore, by utilizing the property of persistence, a quantitative trust model is established for IoT systems based on the past behaviors of the nodes. Then, a trust-based distributed estimation method is proposed to enhance security. Simulation results by considering an IoT system with 50 nodes and 145 edges are also provided to verify our theoretical results.

Driver status monitoring systems are a vital component of smart cars in the future, especially in the era when an increasing amount of time is spent in the vehicle. The heart rate (HR) is one of the most important physiological signals of driver status. To infer HR of drivers, the mainstream of the existing research focused on capturing subtle heartbeat-induced vibration of the torso or leveraged photoplethysmography (PPG) that detects cardiac cycle-related blood volume changes in the microvascular. However, existing approaches rely on dedicated sensors that are expensive and cumbersome to be integrated or are vulnerable to ambient noise. Moreover, their performance on the detection of HR does not guarantee a reliable computation of the HR variability (HRV) measure, which is a more applicable metric for inferring mental and physiological status. The accurate computation HRV measure is based on the precise measurement of the beat-to-beat interval, which can only be accomplished by medical-grade devices that attach electrodes to the body. Considering these existing challenges, we proposed a facial expression-based HRV estimation solution. The rationale is to establish a link between facial expression and heartbeat since both are controlled by the autonomic nervous system. To solve this problem, we developed a tree-based probabilistic fusion neural network approach, which significantly improved HRV estimation performance compared to conventional random forest or neural network methods and the measurements from smartwatches. The proposed solution relies only on commodity camera with a lightweighted algorithm, facilitating its ubiquitous deployment in current and future vehicles. Our experiments are based on 3400 km of driving data from nine drivers collected in a naturalistic field study.

Well known oil recovery factor estimation techniques such as analogy, volumetric calculations, material balance, decline curve analysis, hydrodynamic simulations have certain limitations. Those techniques are time-consuming, require specific data and expert knowledge. Besides, though uncertainty estimation is highly desirable for this problem, the methods above do not include this by default. In this work, we present a data-driven technique for oil recovery factor (limited to water flooding) estimation using reservoir parameters and representative statistics. We apply advanced machine learning methods to historical worldwide oilfields datasets (more than 2000 oil reservoirs). The data-driven model might be used as a general tool for rapid and completely objective estimation of the oil recovery factor. In addition, it includes the ability to work with partial input data and to estimate the prediction interval of the oil recovery factor. We perform the evaluation in terms of accuracy and prediction intervals coverage for several tree-based machine learning techniques in application to the following two cases: (1) using parameters only related to geometry, geology, transport, storage and fluid properties, (2) using an extended set of parameters including development and production data. For both cases model proved itself to be robust and reliable. We conclude that the proposed data-driven approach overcomes several limitations of the traditional methods and is suitable for rapid, reliable and objective estimation of oil recovery factor for hydrocarbon reservoir.

Many methods have been proposed to estimate how much effort is required to build and maintain software. Much of that research tries to recommend a single method - an approach that makes the dubious assumption that one method can handle the diversity of software project data. To address that drawback, we apply a configuration technique called “ROME” (Rapid Optimizing Methods for Estimation), which uses sequential model-based optimization (SMO) to find what configuration settings of effort estimation techniques work best for a particular data set. We test this method using data from 1161 classic waterfall projects and 120 contemporary projects (from GitHub). In terms of magnitude of relative error and standardized accuracy, we find that ROME achieves better performance than the state-of-the-art methods for both classic and contemporary projects. In addition, we conclude that we should not recommend one method for estimation. Rather, it is better to search through a wide range of different methods to find what works best for the local data. To the best of our knowledge, this is the largest effort estimation experiment yet attempted and the only one to test its methods on classic and contemporary projects.

This paper addresses the problem of reconstructing 3D poses of multiple people from a few calibrated camera views. The main challenge of this problem is to find the cross-view correspondences among noisy and incomplete 2D pose predictions. Most previous methods address this challenge by directly reasoning in 3D using a pictorial structure model, which is inefficient due to the huge state space. We propose a fast and robust approach to solve this problem. Our key idea is to use a multi-way matching algorithm to cluster the detected 2D poses in all views. Each resulting cluster encodes 2D poses of the same person across different views and consistent correspondences across the keypoints, from which the 3D pose of each person can be effectively inferred. The proposed convex optimization based multi-way matching algorithm is efficient and robust against missing and false detections, without knowing the number of people in the scene. Moreover, we propose to combine geometric and appearance cues for cross-view matching. Finally, an efficient tracking method is proposed to track the detected 3D poses across the multi-view video. The proposed approach achieves the state-of-the-art performance on the Campus and Shelf datasets, while being efficient for real-time applications.

Direction of arrival (DOA) estimation of wireless signals is demanded in many applications. In addition to classical methods such as MUSIC and ESPRIT, non-linear algorithms such as compressed sensing have become common subjects of study recently. Deep learning or machine learning is also known as a non-linear algorithm and has been applied in various fields. Generally, DOA estimation using deep learning is classified as on-grid estimation. A major problem of on-grid estimation is that the accuracy may be degraded when the DOA is near the boundary. To reduce such estimation errors, we propose a method of combining two DNNs whose grids are offset by one half of the grid size. Simulation results show that our proposal outperforms MUSIC which is a typical off-grid estimation method. Furthermore, it is shown that the DNN specially trained for a close DOA case achieves very high accuracy for that case compared with MUSIC.

Surveying wide areas with only one camera is a typical scenario in surveillance and automotive applications. Ultra wide-angle fisheye cameras employed to that end produce video data with characteristics that differ significantly from conventional rectilinear imagery as obtained by perspective pinhole cameras. Those characteristics are not considered in typical image and video processing algorithms such as motion estimation, where translation is assumed to be the predominant kind of motion. This contribution introduces an adapted technique for use in block-based motion estimation that takes into the account the projection function of fisheye cameras and thus compensates for the non-perspective properties of fisheye videos. By including suitable projections, the translational motion model that would otherwise only hold for perspective material is exploited, leading to improved motion estimation results without altering the source material. In addition, we discuss extensions that allow for a better prediction of the peripheral image areas, where motion estimation falters due to spatial constraints, and further include calibration information to account for lens properties deviating from the theoretical function. Simulations and experiments are conducted on synthetic as well as real-world fisheye video sequences that are part of a data set created in the context of this paper. Average synthetic and real-world gains of 1.45 and 1.51 dB in luminance PSNR are achieved compared against conventional block matching. Furthermore, the proposed fisheye motion estimation method is successfully applied to motion compensated temporal resolution enhancement, where average gains amount to 0.79 and 0.76 dB.

Deteriorating water quality leads to the freshwater biodiversity crisis. The interrelationships among water quality parameters and the relationships between these parameters and taxa groups are complicated in affecting biodiversity. Never-theless, due to the limited types of Internet of Things (IoT) sensors available on the market, a large number of chemical and biological parameters still rely on laboratory tests. With the latest advancement in artificial intelligence and the IoT (AIoT), this technique can be applied to real-time monitoring of water quality, and further conserving biodiversity. In this paper, we conducted a comprehensive literature review on water quality parameters that impact the biodiversity of freshwater and identified the top-10 crucial water quality parameters. Among these parameters, the interrelationships between the IoT measurable parameters and IoT unmeasurable parameters are estimated using a general regression neural network model and a multivariate polynomial regression model based on historical water quality monitoring data. Conventional field water sampling and in-lab experiments, together with the developed IoT-based water quality monitoring system were jointly used to validate the estimation results along an urban river in Hong Kong. The general regression neural network model can successfully distinguish the abnormal increase of parameters against normal situations. For the multivariate polynomial regression model of degree eight, the coefficients of determination results are 0.89, 0.78, 0.87, and 0.81 for NO3-N, BOD5, PO4, and NH3-N, respectively. The effectiveness and efficiency of the proposed systems and models were validated against laboratory results and the overall performance is acceptable with most of the prediction errors smaller than 0.2mg/L, which provides insights into how AIoT techniques can be applied to pollutant discharge monitoring and other water quality regulatory applications for freshwater biodiversity conservation.

Ensuring aggregate food price stability requires a forward-looking assessment of the risk that unexpected deviations in individual food items’ inflation lead to large shocks in the aggregate food price inflation. To do so, we propose using a multivariate GARCH framework in combination with the Euler method to (1) estimate the conditional standard deviation and quantiles of the food price inflation shocks and (2) attribute the total risk to the underlying food items. For the FAO food price index, we find that even though meat inflation systematically has the highest weight in the aggregate index, cereal inflation is the main contributor to the total food price inflation risk over the period 1990–2018. The use of time series models and the Cornish-Fisher expansion make the risk characterization forward-looking and a potentially helpful tool for risk management.

Saturation flow is a fundamental performance measure parameter used for the design of intersections. However, in mixed traffic conditions, the evaluation of saturation flow is more sensitive due to heterogeneous traffic where vehicles are travelling in the same right of way without any lane separation. In such cases, the passenger car unit (PCU) plays a major role in analysing heterogeneity using an equivalent value for each vehicle. This paper presents a novel approach using the optimisation technique to determine PCU values and saturation flow for mixed traffic streams. Field data were collected at six signalised intersections in three Indian cities. The optimisation technique produced reasonable PCU values and an accurate estimation of saturation flow of 1906 PCUs/h. The saturation flow prediction model was developed using the multi-linear regression technique as per Pearson's correlation between saturation flow and vehicle compositional share. The developed saturation flow model was validated using field data from another site and the predicted value was in good agreement with the observed value. Finally, the predictive model was used to analyse the impact of traffic composition on saturation flow in mixed traffic streams.

In this paper, we propose an iterative feature matching framework for self-supervised depth estimation in indoor scenes. Conventional methods usually leverage the structure-from-motion supervision to help the photometric optimization escape from the local minima, which have complex ego-motion and large regions with non-texture or repeated-texture. However, the supervision is limited as the reconstruction is usually sparse. To address this, we propose an iterative feature matching framework called IFMNet to jointly learn depths and search for correspondences. With the predicted depths from the previous iteration, we present an online optimized grid searching algorithm to find more accurate correspondences. Given these new correspondences, we compute the triangulated depths and improve the depth network with adaptive bin-wise online hard example mining. Experimental results on the NYU Depth V2 and SceneNet datasets verify the effectiveness of our approach.

With the increasing attention to the energy crisis and global warming, solar generation has become an important way to use clean solar energy and is playing an increasingly important role. Due to the highly-variable patterns of solar generation, the estimation of prediction intervals is receiving more attention, which is conducive to the safe and stable operation of the power system. In order to further improve the performance of prediction intervals of solar generation, this paper proposes a prediction intervals estimation method for solar generation based on gated recurrent unit (GRU) neural networks and kernel density estimation (KDE). GRU, a commonly used recurrent neural networks, is utilized to obtain the deterministic forecast of solar generation. In addition, according to the characteristics of solar generation, attention mechanism is designed on the GRU prediction model to further improve the prediction performance. Then, the KDE method is used to fit the prediction errors of solar generation obtained by the deterministic forecasting method. In order to verify the effectiveness of the proposed method, we have carried out a large number of experiments on freely available datasets. The experimental results show that the proposed method outperforms competing methods and can generate high-quality prediction intervals.

This paper proposes a method to estimate the structure and motion (SaM) of multiple objects using a monocular camera. Most of the existing methods for SaM estimation are restricted to a single object by constraining the camera and object motions. In this study, SaM estimation is extended for the case of multiple (or simply two) objects and the restrictions on the camera and object motions and the necessity for the information related to the camera velocity are relaxed by using the relative information of the two objects. The relative motion dynamics of partially measurable states of multiple objects are first introduced, based on which the ranges (i.e., depths) of the two objects with respect to the camera, relative velocity between the objects, relative velocity between each object and the camera, and velocities of both the camera and objects are then estimated by using the proposed fuzzy integral sliding mode observer (FISMO). The following cases for the camera and objects-that have either static (stationary), dynamic (moving), or general (i.e., either static or dynamic) motions-are studied: i) two objects are at the same range with respect to the camera; and ii) two objects are at different ranges with respect to the camera. Several situations- in which the ranges of the two objects with respect to the monocular camera, relative velocity between the two objects, and velocities of the camera and objects can be estimated-are described in detail. Simulations and experimental results demonstrate the validity of the proposed method.

With the increasing quantity of private cars, curb parking has evolved into an important approach to mitigate parking pressure in urban cities. While some efforts have been made for the demand analysis of point-of-interest (POI) and pattern analysis of human mobility, which may indirectly reflect the parking situation in urban area, there is a lack of comprehensive models for the parking demand, so as to make a prediction for the road sections without parking lots. In this paper, by focusing on curb parking and designing a systemic framework, named Curb Parking Demand Estimation (CPDE), we model the public parking demand in urban area, w.r.t. parking durations and regional characteristics. Specifically, we use taxi destinations and the distribution of POIs to quantitatively analyze the regional characteristics, designing corresponding features, and propose a K-means-based Least Square (KLS) method to relate parking characteristics, namely, the temporal parking durations and the corresponding demands, with these features. In this way, we effectively avoid the geographical sparsity of road parking sections and can finely estimate parking durations and demands for newly developed districts without parking data. Moreover, we give a strategy, named Parking Types Estimation (PTE), which projects estimated parking durations and demands onto Gaussian Mixture Model (GMM) to accurately measure the distribution of demands over different parking durations for a road section. At last, we conduct experiments on a real-world curb parking dataset in Hefei, a provincial city in China. This dataset contains parking orders of 2016 over the urban area of Hefei. The experimental results validate the effectiveness of our methods, and show that our framework outperforms the state-of-the-art baseline schemes.

Background: The aim of the study was to determine sex and age from hyoid bone morphology on cone-beam computed tomography (CBCT). Materials and methods: The study sample compromised 130 CBCT images. Eight different measurements were performed for each hyoid bone. Fusion conditions were assigned to each side of the hyoid bone separately as; unfused, partially fused, and total fused. Results: The Hyoid length (Hyd-L), width of the left proximal end (LP-W) and stature variables showed sexual dimorphism in all hyoid fusion types (p < 0.05). In young adult age group, a significant difference was found between sex and fusion types (p = 0.025). The body length (B-L) (81.35%) and stature measurements (76.25%) contributed most significantly to sex estimation. Conclusions: Patient’s CBCT scans which enable reproducible and reliable measurements for bone tissues, can be used for forensic procedures. Hyoid bone measurements with CBCT are useful methodology for age and sex estimation in forensic sciences with high accuracy predictive rates.

In this study, an inverse-problem method was applied to estimate the solid concentration in a solid–liquid two-phase flow. An algebraic slip mixture model was introduced to solve the forward problem of solid–liquid convective heat transfer. The time-average conservation equations of mass, momentum, energy, as well as the volume fraction equation were computed in a computational fluid dynamics (CFD) simulation. The solid concentration in the CFD model was controlled using an external program that included the inversion iteration, and an optimal estimation was performed via experimental measurements. Experiments using a fly-ash–water mixture and sand–water mixture with different solid concentrations in a horizontal pipeline were conducted to verify the accuracy of the inverse-problem method. The estimated results were rectified using a method based on the relationship between the estimated results and estimation error; consequently, the accuracy of the corrected inversion results improved significantly. After a verification through experiments, the inverse-problem method was concluded to be feasible for predicting the solid concentration, as the estimation error of the corrected results was within 7% for all experimental samples for a solid concentration of less than 50%. The inverse-problem method is expected to provide accurate predictions of the solid concentration in solid–liquid two-phase flow systems.

Rail surface defect inspection based on machine vision faces challenges against the complex background with interference and severe data imbalance. To meet these challenges, we regard defect detection as a key-point estimation problem and present the attention neural network for rail surface defect detection via CASIoU-guided center-point estimation (CCEANN). CCEANN contains two crucial components. One is the stacked attention Hourglass backbone via cross-stage fusion of multi-scale features (CSFA-Hourglass), in which the convolutional block attention module with variable receptive fields (VRF-CBAM) is introduced, and a two-stage Hourglass structure balancing the network depth and feature fusion plays a key role. Furthermore, the CASIoU-guided center-point estimation head module (CASIoU-CEHM) integrating the delicate coordinate compensation mechanism regresses detection boxes flexibly to adapt to defects' large-scale variation, in which the proposed CASIoU loss, a loss regressing the consistency of Intersection-over-Union (IoU), central-point distance, area ratio, and scale ratio between the targeted defect and the predicted defect, achieves higher regression accuracy than state-of-the-art IoU-based losses. The experiments demonstrate that CCEANN outperforms competitive deep learning-based methods in four surface defect datasets.

Electrochemical mechanisms in lithium-ion batteries (LIBs) pose a significant challenge in deriving models that are highly accurate, have low computational complexity, and enable real-time state and parameter estimation. In this article, we propose a machine learning model as an important building block of a physics-based ANCF-e model that was recently proposed for LIBs. This machine learning model is used to estimate nonlinear potentials, including the open-circuit potential, electrolyte potential, and lithium-intercalation overpotential. Such an estimation is shown to result in a much smaller computational complexity and therefore can enable real-time state and parameter estimation. Three different machine learning architectures are explored, including multilayer perceptron, radial basis function (RBF)-based neural networks, and support vector machines. The training of these machine learning models is carried out using current profiles obtained with an electric vehicle model from driving cycles as inputs and ANCF-e model-based outputs. The underlying ANCF-e model is validated both through a high-fidelity numerical approach, including COMSOL and an experimental test using commercial LIBs. Both validations are carried out under both constant current discharging and dynamic load cycles. The resulting performance using these machine learning models is compared using different metrics, including estimation errors, convergence rates, training time, and computational time. The results indicate that an RBF-based neural network leads to better estimation of the underlying potentials in LIBs and that all machine learning models require a computational time that is 95% smaller than a physics-based approach for this estimation.

INTRODUCTION: Existing automated seizure detection algorithms report sensitivities between 43% and 77% and specificities between 56% and 90%. The algorithms suffer from false alarms when applied to neonatal EEG because of the high degree of nurse handling and rhythmic patting used to soothe neonates. Computer vision technology that quantifies movement in real time could distinguish artifactual motion and improve automated neonatal seizure detection algorithms. METHODS: The authors used video EEG recordings from 43 neonates undergoing monitoring for seizures as part of the NEOLEV2 clinical trial. The Persyst neonatal automated seizure detection algorithm ran in real time during study EEG acquisitions. Computer vision algorithms were applied to extract detailed accounts of artifactual movement of the neonate or people near the neonate though dense optical flow estimation. RESULTS: Using the methods mentioned above, 197 periods of patting activity were identified and quantified, of which 45 generated false-positive automated seizure detection events. A binary patting detection algorithm was trained with a subset of 470 event videos. This supervised detection algorithm was applied to a testing subset of 187 event videos with 8 false-positive events, which resulted in a 24% reduction in false-positive automated seizure detections and a 50% reduction in false-positive events caused by neonatal care patting, while maintaining 11 of 12 true-positive seizure detection events. CONCLUSIONS: This work presents a novel approach to improving automated seizure detection algorithms used during neonatal video EEG monitoring. This artifact detection mechanism can improve the ability of a seizure detector algorithm to distinguish between artifact and true seizure activity.

Crowd counting and crowd density estimation methods are of great significance in the field of public security. Estimating crowd density and counting from single image or video frame has become an essential part of a computer vision system in various scenarios. In this paper, we comprehensively review the recent research advancement on crowd counting and density estimation. First of all, we introduce the background of crowd counting and crowd density estimation. Second, the traditional crowd counting methods are summarized. Third, we focus on reviewing the crowd counting and crowd density methods based on Convolutional Neural Network (CNN) models. Next, we report and discuss the experimental results of a number of typical methods on benchmark datasets. Finally, we present the promising future directions of crowd counting and crowd density.

In recent years, self-supervised paradigms for depth estimation have drawn lots of attention from the community. Promising as they are, in order to achieve wider application and better performance, reasoning about the uncertainty of the estimated depth maps is in great demand. In this paper, we propose a novel uncertainty-aware self-improving framework for self-supervised depth estimation, which could model and exploit uncertainty to boost depth accuracy without any external supervision. In detail, our framework includes two DepthNets–a teacher network and a student network. The teacher network is pre-trained in a self-supervised manner and the predicted depth is leveraged as pseudo-ground-truth to supervise the learning of the student network, leading to better estimation accuracy and the ability of modeling depth uncertainty. Afterward, the estimated uncertainty would be introduced back into the teacher network as the mask, preventing erroneous regions from contributing to network training and depth estimation. In this way, a better pseudo-ground-truth from the teacher network will be used for the next iteration of the student network training. We evaluate our proposed framework on the widely used KITTI dataset and TUM RGB-D dataset with both quantitative and qualitative experiments. The experimental results show that our framework achieves quite satisfactory performance among the self-supervised depth estimation methods. Extensive experiments which combine the student network and feature-based monocular SLAM system (i.e ORB-SLAM3) further demonstrate the benefits of the introduced depth uncertainty.

Height estimation from single images, strictly speaking, is an ill-posed problem. However, recently, it is shown that it is both possible and feasible to learn a mapping from image statistics to height information. In spite of recent efforts in this field, how to learn fine-shape preserving features, such as object boundaries and contours, is still an open issue. In this work, we propose a progressive learning network to estimate height information from single aerial images in a coarse-to-fine manner. In particular, a gated feature aggregation module is introduced to effectively combine low-level and high-level features. The proposed method is validated on three public datasets, including the Vaihingen dataset, the Potsdam dataset, and the DFC2019 dataset. Both quantitative and qualitative experimental results demonstrate that the proposed method can achieve more accurate height estimation from single aerial images, especially with better object boundary and contour preserving capability, than four related height estimation methods.

Background: Anecdotal evidence collected by a community organization suggested high prevalence of lymphatic filariasis (LF) in Sitapur district, Uttar Pradesh. Village volunteers subsequently condu...

Purpose: The purpose of this study was to correlate fetal weight estimation by ultrasound and clinical methods with actual birth weight in KNH.Methodology: This is a prospective comparative study. The design was suitable because it enabled comparison of the predictive value, sensitivity and specificity in estimating fetal weight which is known after birth. Study area was KNH Obstetric wards. The study population was all pregnant women admitted to obstetric wards for elective caesarean delivery and study period was February -March 2016. Data was analysed using SPSS version 20. Categorical variables were presented as proportions in tables and graphs, bars or pie charts). Continuous variable were summarized as means or medians and presented in table form.Results: The findings show that the correlation between actual weight and Ultra Sound estimated weight was significant (r=0.65, p

Class imbalance problem that characterized with the skew distribution towards the majority arises as one challenge in recent years. Many oversampling techniques have been proposed to cope with this problem and some of them combine the oversampling procedure with the clustering algorithm which guaranteeing new synthetic samples being generated in clusters. However far-away samples but with the same minority sub-region are generally clustered into different groups owing to the characteristic of clustering algorithm itself. Therefore, the following oversampling procedure is mostly carried in incomplete minority sub-regions that synthetic samples not well cover the integral minority region. And to our best knowledge, none of existing algorithm is designed to directly estimate minority sub-regions for class imbalance problem. Thus, one new grouping algorithm, named Direction Distribution-based Minority Sub-region Estimation (DDMSE), is firstly proposed. The new algorithm exploits the intuitive observation, that the minority with the same sub-region almost distribute within the same direction when compared to other majority, to estimate minority sub-regions that tactfully ignoring negative impacts brought by the distance factor like in clustering algorithms. Finally, new synthetic samples are generated in those minority sub-regions. And experimental results on real-world datasets show the comparable performance with other state-of-the-art oversampling methods.

Energy is one of the most important requirements for countries and serves as an indicator of the development level of a country. Developments in social, economic and industrial areas have increased the energy needs of countries. Solar energy is one of the most promising renewable energy sources to meet increased energy demands. Photovoltaic (PV) models used to generate electricity from solar energy are among the most popular renewable energy systems. To obtain maximum efficiency from PV models, it is very important to estimate parameters based on current–voltage data, simulate and control PV models for electricity generation. To achieve maximum efficiency for PV models, their parameters should be estimated in an optimum way. Therefore, tree seed algorithm (TSA) developed inspired by nature was proposed in this study. TSA is a meta-heuristic algorithm inspired by the connection between trees and seeds and proposed for solving continuous optimization problems. Parameters of three different PV models (single diode model, double diode model and PV module) were optimized with TSA. Various analyses were performed to prove the success of the proposed method in parameter optimization of PV models. The results were also compared with those of similar studies in the literature. The analysis results indicated that the proposed method yielded very successful, effective and reliable results, and it was concluded that it was a competitive method compared to other methods in the literature.

We present an approach to estimate 3D human pose from a monocular image. The method consists of two steps: it first estimates a 2D pose from an image and then recovers the corresponding 3D pose. This work focuses on the second step. The Graph Convolutional Network (GCN) has recently become the de facto standard for human pose related tasks such as action recognition. However, in this work, we show that it has critical limitations when used for 3D pose estimation due to the inherent weight sharing scheme. The limitations are clearly exposed through a more generic reformulation of GCN, in which both GCN and Fully Connected Network (FCN) are its special cases. In addition, on top of the formulation, we propose Locally Connected Network (LCN) to overcome the limitations of GCN by allocating dedicated rather than shared filters for different joints. We train the LCN network together with the 2D pose estimator such that LCN can be trained to handle inaccurate 2D poses. We evaluate our approach on two benchmarks, and observe that LCN outperforms GCN, FCN and the state-of-the-arts by a large margin. More importantly, it demonstrates stronger cross-dataset generalization ability because of the sparse connections among joints.

Low-density lipoprotein cholesterol (LDL-C) is typically estimated from total cholesterol, high-density lipoprotein cholesterol, and triglycerides. The Friedewald, Martin-Hopkins, and National Institutes of Health equations are widely used but may estimate LDL-C inaccurately in certain patient populations, such as those with HIV. We sought to investigate the utility of machine learning for LDL-C estimation in a large cohort of women with and without HIV.We identified 7397 direct LDL-C measurements (5219 from HIV-infected individuals, 2127 from uninfected controls, and 51 from seroconvertors) from 2414 participants (age 39.4 ± 9.3 years) in the Women's Interagency HIV Study and estimated LDL-C using the Friedewald, Martin-Hopkins, and National Institutes of Health equations. We also optimized 5 machine learning methods [linear regression, random forest, gradient boosting, support vector machine (SVM), and neural network] using 80% of the data (training set). We compared the performance of each method using root mean square error, mean absolute error, and coefficient of determination (R2) in the holdout (20%) set.SVM outperformed all 3 existing equations and other machine learning methods, achieving the lowest root mean square error and mean absolute error, and the highest R2 (11.79 and 7.98 mg/dL, 0.87, respectively, compared with those obtained using the Friedewald equation: 12.45 and 9.14 mg/dL, 0.87). SVM performance remained superior in subgroups with and without HIV, with nonfasting measurements, in LDL70 mg/dL and triglycerides400 mg/dL.In this proof-of-concept study, SVM is a robust method that predicts directly measured LDL-C more accurately than clinically used methods in women with and without HIV. Further studies should explore the utility in broader populations.

Computer-aided translation tools based on translation memories are widely used to assist professional translators. A translation memory (TM) consists of a set of translation units (TU) made up of source- and target-language segment pairs. For the translation of a new source segment s', these tools search the TM and retrieve the TUs (s,t) whose source segments are more similar to s'. The translator then chooses a TU and edit the target segment t to turn it into an adequate translation of s'. Fuzzy-match repair (FMR) techniques can be used to automatically modify the parts of t that need to be edited. We describe a language-independent FMR method that first uses machine translation to generate, given s' and (s,t), a set of candidate fuzzy-match repaired segments, and then chooses the best one by estimating their quality. An evaluation on three different language pairs shows that the selected candidate is a good approximation to the best (oracle) candidate produced and is closer to reference translations than machine-translated segments and unrepaired fuzzy matches (t). In addition, a single quality estimation model trained on a mix of data from all the languages performs well on any of the languages used. This work was supported by the Spanish Government through the EFFORTUNE project [TIN-2015-69632-R].

The effect of tertiary flow length on asphalt concrete (AC) flow number (FN) has been studied in this paper. The standard FN test (AASHTO T378) designated by American Association of State Highway and Transportation Officials (AASHTO) is currently among the widely adopted test for assessing the rutting performance of AC mixtures. The standard adopts the Fracken model (FM) for fitting the permanent deformation curve (PDC), prior to FN estimation. The FN was observed to vary as tertiary flow progresses. The FM, along with other PDC models (MFM-1 and MFM-2) obtained by modifying the FM, was utilized to analyze and minimized this variation. Variation of FN resulted because of mathematization of the PDC data. Instead of representing the actual initiating point of permanent shear deformation of the strain data, estimated values of FN were observed to indicate the inflection points of the fitted parent curve. As per standard FN range suggested by AASHTO T378, the observed variation in FN leads to the situation, where a single asphalt mix specimen can be regarded as appropriate for two different traffic levels, depending on the tertiary flow point at which the test is terminated. Supplementary steps in the FN estimation methods were proposed for refinement of FN values. For the refinement and further standardization of FN value and FN test respectively, FN:T (flow number to test duration ratio) has been recognized as the key and most simple solution. Two potential options for employing FN:T in the estimation of FN have also been highlighted. Several important correlations have been reported herein as well. Comparatively, MFM-1 was found to be more robust in FN:T curve utilization than MFM-2 and FM.

In the recent years, the advances in mobile technology have brought an exorbitant change in daily lifestyle of individuals. Smartphones/mobile devices are rampant in all aspects of human life. This has led to an extreme demand for developing software that runs on mobile devices. The developers have to keep up with this high demand and deliver high-quality app on time and within budget. For this, estimation of development and testing of apps play a pivotal role. In this paper, a Systematic Literature Review (SLR) is conducted to highlight development and testing estimation process for software/application. The goal of the present literature survey is to identify and compare existing test estimation techniques for traditional software (desktop/laptop) and for mobile software/application. The characteristics that make mobile software/application different from traditional software are identified in this literature survey. Further, the trend for developing the software is towards agile, thus this study also presents and compares estimation techniques used in agile software development for mobile applications. The analysis of literature review suggests filling a research gap to present formal models for estimating mobile application considering specific characteristics of mobile software.

In this study, statistical analysis is performed in order to characterize wind speeds distribution according to different samples randomly drawn from wind speed data collected. The purpose of this study is to assess how random sampling influences the estimation quality of the shape (k) and scale (c) parameters of a Weibull distribution function. Five stations were chosen in West Africa for the study, namely: Accra Kotoka, Cotonou Cadjehoun, Kano Mallam Aminu, Lomé Tokoin and Ouagadougou airport. We used the energy factor method (EPF) to compute shape and scale parameters. Statistical indicators used to assess estimation accuracy are the root mean square error (RMSE) and relative percentage error (RPE). Study results show that good accuracy in Weibull parameters and power density estimation is obtained with sampled wind speed data of 30% for Accra, 20% for Cotonou, 80% for Kano, 20% for Lomé, and 20% for Ouagadougou site. This study showed that for wind potential assessing at a site, wind speed data random sampling is sufficient to calculate wind power density. This is very useful in wind energy exploitation development.

Nowcasts of strong convective precipitation and radar-based quantitative precipitation estimations have always been hot yet challenging issues in meteorological sciences. Data-driven machine learning, especially deep learning, provides a new technical approach for the quantitative estimation and forecasting of precipitation. A high-quality, large-sample, and labeled training dataset is critical for the successful application of machine-learning technology to a specific field. The present study develops a benchmark dataset that can be applied to machine learning for minute-scale quantitative precipitation estimation and forecasting (QpefBD), containing 231,978 samples of 3,185 heavy precipitation events that occurred from April to October during the years 2016 to 2018 in six provinces in central and eastern China. Each individual sample consists of eight products of weather radars at 6-min intervals within the time window of the corresponding event and products of 27 physical quantities at hourly intervals that describe the atmospheric dynamic and thermodynamic conditions. Two data labels, i.e., ground precipitation intensity and areal coverage of heavy precipitation at 6-minute intervals, are also included. The present study describes the basic components of the dataset and data processing and provides metrics for the evaluation of model performance on precipitation estimation and forecasting. Based on these evaluation metrics, some simple, commonly used methods are applied to evaluate precipitation estimates and forecasts. The results can serve as the benchmark reference for the performance evaluation of machine learning models using this dataset. This paper also gives some suggestions and scenarios of the QpefBD application. We believe that the application of this benchmark dataset will promote interdisciplinary collaboration between meteorological sciences and artificial intelligence sciences, providing a new way for the identification and forecast of heavy precipitation.

This paper presents a Bayesian framework for estimating a Probabilistic Linear Discriminant Analysis (PLDA) model in the presence of noisy labels. True class labels are interpreted as latent random variables, which are transmitted through a noisy channel, and received as observed speaker labels. The labeling process is modeled as a Discrete Memoryless Channel (DMC). PLDA hyperparameters are interpreted as random variables, and their joint posterior distribution is derived using mean-field Variational Bayes, allowing maximum a posteriori (MAP) estimates of the PLDA model parameters to be determined. The proposed solution, referred to as VB-MAP, is presented as a general framework, but is studied in the context of speaker verification, and a variety of use cases are discussed. Specifically, VB-MAP can be used for PLDA estimation with unreliable labels, unsupervised PLDA estimation, and to infer the reliability of a PLDA training set. Experimental results show the proposed approach to provide significant performance improvements on a variety of NIST Speaker Recognition Evaluation (SRE) tasks, both for data sets with simulated mislabels, and for data sets with naturally occurring missing or unreliable labels.

Numerical predictive modeling is widely used in different application domains. Although many modeling techniques have been proposed, and a number of different aggregate accuracy metrics exist for evaluating the overall performance of predictive models, other important aspects, such as the reliability (or confidence and uncertainty) of individual predictions, have been underexplored. We propose to use estimated absolute prediction error as the indicator of individual prediction reliability, which has the benefits of being intuitive and providing highly interpretable information to decision makers, as well as allowing for more precise evaluation of reliability estimation quality. As importantly, the proposed reliability indicator allows the reframing of reliability estimation itself as a canonical numeric prediction problem, which makes the proposed approach general-purpose (i.e., it can work in conjunction with any outcome prediction model), alleviates the need for distributional assumptions, and enables the use of advanced, state-of-the-art machine learning techniques to learn individual prediction reliability patterns directly from data. Extensive experimental results on multiple real-world data sets show that the proposed machine learning-based approach can significantly improve individual prediction reliability estimation as compared with a number of baselines from prior work, especially in more complex predictive scenarios.

Purpose To accelerate non-rigid motion corrected coronary MR angiography (CMRA) reconstruction by developing a deep learning based non-rigid motion estimation network and combining this with an efficient implementation of the undersampled motion corrected reconstruction. Methods Undersampled and respiratory motion corrected CMRA with overall short scans of 5 to 10 min have been recently proposed. However, image reconstruction with this approach remains lengthy, since it relies on several non-rigid image registrations to estimate the respiratory motion and on a subsequent iterative optimization to correct for motion during the undersampled reconstruction. Here we introduce a self-supervised diffeomorphic non-rigid respiratory motion estimation network, DiRespME-net, to speed up respiratory motion estimation. We couple this with an efficient GPU-based implementation of the subsequent motion-corrected iterative reconstruction. DiRespME-net is based on a U-Net architecture, and is trained in a self-supervised fashion, with a loss enforcing image similarity and spatial smoothness of the motion fields. Motion predicted by DiRespME-net was used for GPU-based motion-corrected CMRA in 12 test subjects and final images were compared to those produced by state-of-the-art reconstruction. Vessel sharpness and visible length of the right coronary artery (RCA) and the left anterior descending (LAD) coronary artery were used as metrics of image quality for comparison. Results No statistically significant difference in image quality was found between images reconstructed with the proposed approach (MC:DiRespME-net) and a motion-corrected reconstruction using cubic B-splines (MC:Nifty-reg). Visible vessel length was not significantly different between methods (RCA: MC:Nifty-reg 5.7 ± 1.7 cm vs MC:DiRespME-net 5.8 ± 1.7 cm, P = 0.32; LAD: MC:Nifty-reg 7.0 ± 2.6 cm vs MC:DiRespME-net 6.9 ± 2.7 cm, P = 0.81). Similarly, no statistically significant difference between methods was observed in terms of vessel sharpness (RCA: MC:Nifty-reg 60.3 ± 7.2% vs MC:DiRespME-net 61.0 ± 6.8%, P = 0.19; LAD: MC:Nifty-reg 57.4 ± 7.9% vs MC:DiRespME-net 58.1 ± 7.5%, P = 0.27). The proposed approach achieved a 50-fold reduction in computation time, resulting in a total reconstruction time of approximately 20 s. Conclusions The proposed self-supervised learning-based motion corrected reconstruction enables fast motion-corrected CMRA image reconstruction, holding promise for integration in clinical routine.

Precession of the cone-shaped warhead is the typical micromotion of ballistic targets, and three-dimensional (3D) inverse synthetic aperture radar (ISAR) imaging of the precession warhead is of great significance for ballistic target identification. This letter proposes a novel method for the precession parameter estimation and the 3D ISAR imaging of the cone-shaped target. First, the imaging model of the precession cone-shaped target is developed, and the relationship between the radial-range curve on the range–slow-time plane and the precession parameters is derived. Subsequently, the wideband measurements from two-aspect radars are jointly processed to solve the target precession parameters implicitly involved in the radial-range curve of the scatterer at the cone top. Finally, based on the estimation of precession parameters, the 3D images of the target can be reconstructed, which reveal the structure characteristics and spatial attitude of the target. Simulation results verify the effectiveness of the proposed method.

We present a method to estimate and quantify geometric parameters of a buried object in near-field microwave images. These parameters are shape, location, size, and orientation. Unlike the parameter estimation based on a conventional ground-penetrating radar image, a 3-D well-focused coherent image is generated. Instead of choosing a z-slice of the 3-D image, both 2-D projection and shadow projection images are used to perform the parameter estimation. The shape of the buried object is determined based on both x- and y-polarized images. Other geometric parameters, such as location, size, and orientation, are estimated by the method of spatial moments. Both numerical and experimental data of an object buried in a dielectric slab are used to demonstrate the efficacy of the proposed geometric parameter estimation algorithm.

Global Software Development (GSD) is becoming increasingly prevalent, with software development teams being distributed around the world and working in collaboration with partner companies despite geographic and time differences. The main advantage of GSD which makes it attractive is the greater availability of human resources at lower costs. However, there are several disadvantages which are caused by the distance which separates the development teams. Coordination and communication become more difficult when the software development teams are located in different places, resulting in hidden costs involved in this process. As such, the effort estimation models used for collocated software development are inadequate for estimation in GSD. Thus, effort estimation in GSD is becoming an important area of research. Many researchers have focused on effort estimation in GSD over the last decade. This paper presents the findings of a systematic review of the related literature by summarizing the hidden costs in GSD, and discussing the open research issues in effort estimation in GSD.

In classical statistics, there is a close link between null hypothesis significance testing (NHST) and parameter estimation via confidence intervals. However, for the Bayesian counterpart, a link between null hypothesis Bayesian testing (NHBT) and Bayesian estimation via a posterior distribution is less straightforward, but does exist, and has recently been reiterated by Rouder, Haaf, and Vandekerckhove (2018). It hinges on a combination of a point mass probability and a probability density function as prior (denoted as the spike-and-slab prior). In the present paper, it is first carefully explained how the spike-and-slab prior is defined, and how results can be derived for which proofs were not given in Rouder, Haaf, and Vandekerckhove (2018). Next, it is shown that this spike-and-slab prior can be approximated by a pure probability density function with a rectangular peak around the center towering highly above the remainder of the density function. Finally, we will indicate how this ‘hill-and-chimney’ prior may in turn be approximated by fully continuous priors. In this way, it is shown that NHBT results can be approximated well by results from estimation using a strongly peaked prior, and it is noted that the estimation itself offers more than merely the posterior odds on which NHBT is based. Thus, it complies with the strong APA requirement of not just mentioning testing results but also offering effect size information. It also offers a transparent perspective on the NHBT approach employing a prior with a strong peak around the chosen point null hypothesis value.

The rapid growth of battery electric vehicles (BEVs) in the automotive field has led to the need for improving their drivetrain performance, mainly focusing on the extension of the battery operating range. However, the majority of performed technical assessments only consider battery state-of-charge (SoC) and depth of discharge, while neglecting the effects on lifetime and failure probability of power electronic components, more specifically the emerging wide bandgap (WBG)-based technologies. Toward fulfilling this gap, the EV market demands lifetime estimation carried out under real-life mission profile to confirm efficiency and reliable operation of EV power electronics for an extended range meeting the EVs lifetime requirements. In this regard, this study proposes a versatile experimental device-under-test setup to investigate a novel stepwise holistic system-level lifetime estimation approach for BEV drivetrains equipped with SiC interleaved bidirectional HV dc/dc converter (IBC). To this end, three different real-life mission profile use-cases are investigated in this article and they provide systematic stress-based lifetime estimation, statistical analyses, and validations in comparison. The study outcome highlights realistic information related to significant impacts of operation range and battery SoC features on the IBC lifetime from all aspects.