Arbitrary Regions Research Articles

Image Inpainting by End-to-End Cascaded Refinement With Mask Awareness.

Inpainting arbitrary missing regions is challenging because learning valid features for various masked regions is nontrivial. Though U-shaped encoder-decoder frameworks have been witnessed to be successful, most of them share a common drawback of mask unawareness in feature extraction because all convolution windows (or regions), including those with various shapes of missing pixels, are treated equally and filtered with fixed learned kernels. To this end, we propose our novel mask-aware inpainting solution. Firstly, a Mask-Aware Dynamic Filtering (MADF) module is designed to effectively learn multi-scale features for missing regions in the encoding phase. Specifically, filters for each convolution window are generated from features of the corresponding region of the mask. The second fold of mask awareness is achieved by adopting Point-wise Normalization (PN) in our decoding phase, considering that statistical natures of features at masked points differentiate from those of unmasked points. The proposed PN can tackle this issue by dynamically assigning point-wise scaling factor and bias. Lastly, our model is designed to be an end-to-end cascaded refinement one. Supervision information such as reconstruction loss, perceptual loss and total variation loss is incrementally leveraged to boost the inpainting results from coarse to fine. Effectiveness of the proposed framework is validated both quantitatively and qualitatively via extensive experiments on three public datasets including Places2, CelebA and Paris StreetView.

Fuzzy Adaptive Control for Fractional Nonlinear Systems with External Disturbances and Unknown Control Directions

In this paper, the problem of fuzzy adaptive control of unknown nonlinear fractional-order systems with external disturbances and unknown control directions is studied. We exploit a decomposition of the control gain matrix into a symmetric positive-definite matrix, a diagonal matrix with diagonal entries + 1 or 1 , and a unity upper triangular matrix. Fuzzy logic systems are used for estimating the unknown nonlinear functions. Based on the fractional Lyapunov direct method and some proposed lemmas, a novel fuzzy adaptive controller is designed. The proposed method can guarantee that all the signals in the closed-loop systems remain bounded and the tracking errors converge to an arbitrary small region of the origin. In addition, for updating the parameters of the fuzzy system, fractional-order adaptations laws are proposed. Lastly, an illustrative example is given to demonstrate the effectiveness of the proposed results.

An autopilot for energy models – Automatic generation of renewable supply curves, hourly capacity factors and hourly synthetic electricity demand for arbitrary world regions

Energy system models are increasingly being used to explore scenarios with large shares of variable renewables. This requires input data of high spatial and temporal resolution and places a considerable preprocessing burden on the modeling team. Here we present a new code set with an open source license for automatic generation of input data for large-scale energy system models for arbitrary regions of the world, including sub-national regions, along with an associated generic capacity expansion model of the electricity system. We use ECMWF ERA5 global reanalysis data along with other public geospatial datasets to generate detailed supply curves and hourly capacity factors for solar photovoltaic power, concentrated solar power, onshore and offshore wind power, and existing and future hydropower. Further, we use a machine learning approach to generate synthetic hourly electricity demand series that describe current demand, which we extend to future years using regional SSP scenarios. Finally, our code set automatically generates costs and losses for HVDC interconnections between neighboring regions. The usefulness of our approach is demonstrated by several different case studies based on input data generated by our code. We show that our model runs of a future European electricity system with high share of renewables are in line with results from more detailed models, despite our use of global datasets and synthetic demand.

An error compensation method based on machine vision for laser-processing systems with galvanometers

Laser-processing system plays an important role in modern processing. To improve the precision of a galvanometer laser-processing system, a novel compensation method based on machine vision is proposed. First, the various errors existing in the scanning process of the galvanometer system and the influence on the accuracy of the scanning system were analyzed. Second, processing a specific grid within the processing range of the laser-processing system, the machine vision method was employed to extract the skeleton image and obtain the sub-pixel coordinates of each grid corner point. By comparing them with their theoretical positions, the position and error of all corner points of the mesh were obtained. Finally, the position and error of the four corner points in the arbitrary grid region were analyzed by the weighted interpolation method, and the error value and compensation value of any point in the grid were derived, and an error compensation model was established to compensate the scanning system. To verify the effectiveness of the proposed compensation method, a compensation experiment was conducted, and the workpiece machining accuracy after compensation was improved. The method is simple in operation, offers good compensation effect, and has important theoretical significance and practical value for improving the precision of laser-processing systems.

The plasma characteristics of femtosecond-nanosecond dual-pulse laser ablated soil

We report on the observation of dual-pulses laser-produced plasmas (DP-LPPs), following a femtosecond Ti: sapphire pulse laser ablate soil generated asubstantial numberof soil aerosol particle clusters of soil target and then a nanosecond Nd:YAG pulse laser re-ablated the soil aerosol particles clusters. The temporal evolution of the fs laser-produced plasmas (fs-LPPs) highlight a different in plasma plume expansion dynamics and particles clusters distributions where, the initial position and velocity of plasma plume expansion are 0.79 mm and 4.8 km/s, the terminal position is 2.2 mm and the velocity gradually become slower and slower, in the delay time range of 150 ns to 2.1 μs. There are three distribution states contains a columnar particle fluid flow region, “Mushroom cloud” fluid flow region, and arbitrary shape fluid flow region of the particles scattering images of different inter-pulse time at 1, 5, 9, 10, 30, 50, 200, 400 and 800 μs. Based on the Boltzmann plot method and Stark broadening, the plasma temperature and electron density is also investigated, the results shows that DP-LPPs optical emission spectrum (OES) has a narrow linear broadening and enhanced spectral intensity, associates with the higher plasma temperature and lower electron density compare with the single fs-LPPs.

EasyQuake: Putting Machine Learning to Work for Your Regional Seismic Network or Local Earthquake Study

AbstractWe developed a Python package—easyQuake—that consists of a flexible set of tools for detecting and locating earthquakes from International Federation of Digital Seismograph Networks-collected or field-collected seismograms. The package leverages a machine-learning driven phase picker, coupled with an associator, to produce a Quake Markup Language (QuakeML) style catalog complete with magnitudes and P-wave polarity determinations. We describe how nightly computations on day-long seismograms identify lower-magnitude candidate events that were otherwise missed due to cultural noise and how those events are incorporated into the Oklahoma Geological Survey statewide network upon analyst manual review. We discuss applications for the package, including earthquake detection for regional networks and microseismicity studies in arbitrary user-defined regions. Because the fundamentals of the package are scale invariant, it has wide application to seismological earthquake analysis from regional to local arrays and has great potential for identifying early aftershocks that are otherwise missed. The package is fast and reliable; the computations are relatively efficient across a range of hardware, and we have encountered very few (∼1%) false positive event detections for the Oklahoma case study. The utility and novelty of the package is the turnkey earthquake analysis with QuakeML file output, which can be dropped directly into existing real-time earthquake analysis systems. We have designed the functions to be quite modular so that a user could replace the provided picker or associator with one of their choosing. The Python package is open source and development continues.

Residual Dual Scale Scene Text Spotting by Fusing Bottom-Up and Top-Down Processing

Existing methods for arbitrary shaped text spotting can be divided into two categories: bottom-up methods detect and recognize local areas of text, and then group them into text lines or words; top-down methods detect text regions of interest, then apply polygon fitting and text recognition to the detected regions. In this paper, we analyze the advantages and disadvantages of these two methods, and propose a novel text spotter by fusing bottom-up and top-down processing. To detect text of arbitrary shapes, we employ a bottom-up detector to describe text with a series of rotated squares, and design a top-down detector to represent the region of interest with a minimum enclosing rotated rectangle. Then the text boundary is determined by fusing the outputs of two detectors. To connect arbitrary shaped text detection and recognition, we propose a differentiable operator named RoISlide, which can extract features for arbitrary text regions from whole image feature maps. Based on the extracted features through RoISlide, a CNN and CTC based text recognizer is introduced to make the framework free from character-level annotations. To improve the robustness against scale variance, we further propose a residual dual scale spotting mechanism, where two spotters work on different feature levels, and the high-level spotter is based on residuals of the low-level spotter. Our method has achieved state-of-the-art performance on four English datasets and one Chinese dataset, including both arbitrary shaped and oriented texts. We also provide abundant ablation experiments to analyze how the key components affect the performance.

Time-dependent variational principle for mixed matrix product states in the thermodynamic limit

We describe a time evolution algorithm for quantum spin chains whose Hamiltonians are composed of an infinite uniform left and right bulk part, and an arbitrary finite region in between. The left and right bulk parts are allowed to be different from each other. The algorithm is based on the time-dependent variational principle (TDVP) of matrix product states. It is inversion-free and very simple to adapt from an existing TDVP code for finite systems. The importance of working in the projective Hilbert space is highlighted. We study the quantum Ising model as a benchmark and an illustrative example. The spread of information after a local quench is studied in both the ballistic and the diffusive case. We also offer a derivation of TDVP directly from symplectic geometry.

Experimental demonstration of an arbitrary shape dc electric concentrator

Coordinate transformation (CT) theory has shown great potentials in manipulating both time-varying and static fields for different physics ranging from electromagnetism and acoustics to electrostatic and thermal science. Nevertheless, as inhomogeneous and anisotropic materials are required to be realized for the implementation of CT-based devices, the applicability of this method is restricted due to difficulties in the fabrication process. In this paper, based on transformation electrostatic (TE) methodology, the design principle of an arbitrary shape dc electric concentrator is established which yields the enhancement of static electric fields in a predefined region with only one homogeneous conductivity, named as dc null medium (DNM). It is shown that one constant DNM is sufficient for localizing steady electric current in any arbitrary shape region, which in turn obviates the tedious mathematical calculations that conventional methods suffer from. In other words, the same DNM can be used for different concentrators regardless of their cross-section geometries, which makes the presented approach suitable for scenarios where reconfigurability is of utmost importance. Several numerical simulations are performed in order to demonstrate the capability of the proposed dc electric concentrator in localizing steady electric fields into the desired region. Moreover, by utilizing the analogy between electrically conducting materials and resistor networks, the attained DNM is realized with low-cost resistors and then exploited for fabricating a square shape dc electric concentrator on a printed circuit board (PCB). It is demonstrated that the measurement results agree well with the theoretical predictions and numerical simulations, which corroborate the effectiveness of the propounded method. The presented idea of this paper could find applications in scenarios where highly confined electric fields/currents are of critical importance such as electronic skin devices and electrical impedance tomography.

Flow Instability Investigation at Supercritical Pressures in Parallel Channels by Computational Fluid Dynamics Adopting Low- and High-Power Boundaries

Abstract Supercritical water-cooled reactor (SCWR), which is considered as the logical extension of existing light water reactors (LWRs) (pressurized water reactor and boiling water reactor (BWR)), has the potential of increasing the efficiency of power generation to 45% compared to 33% of that of LWRs. But without the challenges of heat transfer and hydrodynamics, and reactor core design materials due to supercritical flow instability which is associated with sharp variation in fluid properties near the vicinity of the pseudo-critical temperature. Supercritical flow instability therefore needs to be addressed ahead of the deployment and operation of SCWR in the near future. The main purpose of this study is to carry out flow instability analysis in parallel channels with supercritical water. The study also aims at examining the capability of using three-dimensional (3D) simulation of turbulent flow in arbitrary regions computational continuum mechanics C++ based code (3D STAR-CCM+ CFD code) to predict flow oscillation amplitude and periods, and instability power boundaries at low-power boundary (LPB) and at high-power boundary (HPB). Parameters considered in the investigation include mass flowrate, system pressure, and gravity. Two different threshold power instability boundaries were obtained from the study. These instability power boundaries include lower threshold where stability of the parallel channel system decreases with increasing coolant inlet temperature, and upper threshold where stability of the parallel channel system increases with increasing coolant inlet temperature. From the results of the investigation, it can be found that: (1) for LPB at 23 MPa, only lower threshold was obtained as flow instability power boundary; and for HPB at 23 MPa, both lower and upper thresholds were obtained as flow instability power boundaries. The numerical findings quite well agree with the experimental findings at 23 MPa for both LPB and HPB; (2) only lower threshold was obtained as flow instability power boundary at both 23 MPa and 25 MPa for LPB. For HPB, both lower and upper thresholds were obtained as flow instability power boundaries at both 23 MPa and 25 MPa; (3) only lower threshold was obtained as flow instability power boundary for the parallel channel system with or without gravity influence for LPB. For HPB, both lower and upper threshold flow instability power boundaries were obtained for the parallel channel system with gravity influence, but only lower threshold flow instability power boundary was obtained for system without gravity influence; (4) only lower threshold was obtained as flow instability power boundary at system mass flowrates of 125 kg/h and 145 kg/h for LPB. For HPB, both lower and upper threshold flow instability power boundaries were obtained for system mass flowrate of 125 kg/h, but only lower threshold flow instability power boundary was obtained for system mass flowrate of 145 kg/h. For both LPB and HPB, the numerical findings agree quite well with the experimental results for a system operated at 125 kg/h and 145 kg/h; (5) the investigated parameters such as mass flowrate, pressure, and gravity have significant effects on amplitude of mass flow oscillation, but have little effects on the period of mass flow oscillation for both LPB and HPB. Results from the numerical simulation were compared with the results from the experiment for both LPB and HPB. The numerical amplitude results obtained were far less than the amplitude results obtained from the experiment. But there was no significant difference between the oscillation periods obtained from both the numerical simulation and experiment. (6) Flow instability studies including predicting flow oscillation amplitude and periods, and instability power boundaries could be carried out using 3D STAR-CCM+ CFD code. The effects of heating structures on flow instability results have not been considered in this study. Previous studies have shown that including heating structures in geometrical models for numerical studies may have effects on flow instability results. More experimental studies are needed for validation of similar numerical studies carried out at supercritical pressures using various numerical tools.

Super-resolution and denoising of 4D-Flow MRI using physics-Informed deep neural nets

Background and Objective: Time resolved three-dimensional phase contrast magnetic resonance imaging (4D-Flow MRI) has been used to non-invasively measure blood velocities in the human vascular system. However, issues such as low spatio-temporal resolution, acquisition noise, velocity aliasing, and phase-offset artifacts have hampered its clinical application. In this research, we developed a purely data-driven method for super-resolution and denoising of 4D-Flow MRI.Methods: The flow velocities, pressure, and the MRI image magnitude are modeled as a patient-specific deep neural net (DNN). For training, 4D-Flow MRI images in the complex Cartesian space are used to impose data-fidelity. Physics of fluid flow is imposed through regularization. Creative loss function terms have been introduced to handle noise and super-resolution. The trained patient-specific DNN can be sampled to generate noise-free high-resolution flow images. The proposed method has been implemented using the TensorFlow DNN library and tested on numerical phantoms and validated in-vitro using high-resolution particle image velocitmetry (PIV) and 4D-Flow MRI experiments on transparent models subjected to pulsatile flow conditions.Results: In case of numerical phantoms, we were able to increase spatial resolution by a factor of 100 and temporal resolution by a factor of 5 compared to the simulated 4D-Flow MRI. There is an order of magnitude reduction of velocity normalized root mean square error (vNRMSE). In case of the in-vitro validation tests with PIV as reference, there is similar improvement in spatio-temporal resolution. Although the vNRMSE is reduced by 50%, the method is unable to negate a systematic bias with respect to the reference PIV that is introduced by the 4D-Flow MRI measurement.Conclusions: This work has demonstrated the feasibility of using the readily available machinery of deep learning to enhance 4D-Flow MRI using a purely data-driven method. Unlike current state-of-the-art methods, the proposed method is agnostic to geometry and boundary conditions and therefore eliminates the need for tedious tasks such as accurate image segmentation for geometry, image registration, and estimation of boundary flow conditions. Arbitrary regions of interest can be selected for processing. This work will lead to user-friendly analysis tools that will enable quantitative hemodynamic analysis of vascular diseases in a clinical setting.

Direct calculation of mutual information of distant regions

We consider the (Rényi) mutual information, {I}^{(n)}left(A,Bright)={S}_A^{(n)}+{S}_B^{(n)}-{S}_{Acup B}^{(n)} , of distant compact spatial regions A and B in the vacuum state of a free scalar field. The distance r between A and B is much greater than their sizes RA,B. It is known that {I}^{(n)}left(A,Bright)sim {C}_{AB}^{(n)}{leftlangle 0left|phi (r)phi (0)0right|rightrangle}^2 . We obtain the direct expression of {C}_{AB}^{(n)} for arbitrary regions A and B. We perform the analytical continuation of n and obtain the mutual information. The direct expression is useful for the numerical computation. By using the direct expression, we can compute directly I(A, B) without computing SA, SB and SA∪B respectively, so it reduces significantly the amount of computation.

Orthorectified Polar Format Algorithm for Generalized Spotlight SAR Imaging With DEM

In conventional polar format algorithm (PFA), the effective imaging scene is bounded to a limited region near the reference point unless postprocessing is utilized. In a previous paper, refocusing and zoom-in PFA (RZPFA) for curvilinear spotlight SAR imaging were proposed to produce a refocused image for an arbitrary region of interest (ROI) with constant elevation. However, for certain applications, the residual distortion and defocus caused by rugged terrain could not be ignored. In this article, RZPFA is adapted to incorporate the known digital elevation model (DEM) into the imaging process, which is named as orthorectified PFA (OPFA). With just little additional computations than RZPFA, OPFA can realize georeferenced orthorectified imaging via a nonuniform fast Fourier transform of type 3 (NuFFT-3) without the need of postprocessing. The quantitative metrics for the residual DEM distortion and residual DEM defocus were also derived to determine the effective imaging extent of OPFA. Within the effective extent, the proposed OPFA can obtain an orthorectified image efficiently, and the image has a very high quality comparable to backprojection (BP). The imaging results of measured echo and DEM data demonstrated the effectiveness of the proposed algorithm.

Design of freeforms to uniformly illuminate polygonal targets from extended sources via edge ray mapping.

In this article, we consider the design of a compact freeform optical surface that uniformly irradiates an arbitrary convex polygonal region from an extended light source, while controlling spill. This problem has attracted a large body of literature that has primarily covered highly symmetric special cases or cases where the solution is approximated by a zero-étendue design based on a point source. Practical versions of this illumination design problem will likely feature large asymmetric light-emitting diodes, compact lenses, and irregular targets on angled projection surfaces. For these settings, we develop a solution method based on an edge ray mapping that routes maximally off-axis rays from the edges of the source through the edge of the optic to the edges of the target polygon. This determines the sag and normals along the boundary of the freeform surface. A "spill-free" surface is then interpolated from the boundary information and optimized to uniformize the irradiance, while preserving the polygonal boundary. Highly uniform irradiances (relative standard deviation <.01) can be attained with good control of spill, even when the exit surface is less than three source diameters from the embedded source.

Target-oriented full-waveform inversion using Marchenko redatumed wavefields

SUMMARY Full-waveform inversion (FWI) has demonstrated increasing success in estimating medium properties, but its computational cost still poses challenges in moving towards high-resolution imaging of targets at depth. Here, we propose a target-oriented FWI method that inverts for the medium parameters confined within an arbitrary region of interest. Our method is novel in terms of both local wavefield modelling and data redatuming, in order to build a target-oriented objective function which is sensitive to the target medium only without further assumptions about the medium outside. Based on the convolution-type representation theorem, our local forward modelling operator propagates wavefields within the target medium only while providing full acoustic coupling between the target medium and the surrounding geology. A key requirement of our local FWI method is that the subsurface wavefields surrounding and inside the target be as accurate as possible. As such, the subsurface wavefields are retrieved by the Marchenko method, which can redatum the single-sided surface reflection data to the target zone while preserving both primary and multiple reflections, with minimal a priori knowledge of the full-domain medium. Given a sufficiently accurate initial velocity macromodel, our numerical examples show that our local FWI method resolves the reservoir zone of a 2-D Barrett Unconventional P-wave velocity model much more efficiently than the conventional full-domain FWI without significantly sacrificing accuracy. Our method may further enable FWI approaches to high-resolution imaging of subsurface targets.

Aging-Related Hypometabolism in the Anterior Cingulate Cortex of Cognitively Intact, Amyloid-Negative Seniors at Rest Mediates the Relationship between Age and Executive Function but Not Memory.

The anterior cingulate cortex (ACC) shows the most aging-related brain metabolic dysfunction that correlates with decreasing executive processing in otherwise healthy, cognitively intact volunteers. Here, data from ADNI are used to elucidate potential pathophysiological mechanisms involved in cognitive aging, that is, age-related decline in cognitive performance in the absence of known neurodegenerative disease. Amyloid-negative volunteers showed statistically significant mediation of ACC metabolism in the relationship between age and verbal fluency. A nonlinguistic task of executive function, Trails B, showed also negative correlation between performance and age, albeit weaker, but was not significant in the mediation analysis. Recall of story items, minimizing attentional demands compared with learning of word lists, did not correlate with age. ADNI subjects selected for low vascular risks also showed correlation between age and declining ACC metabolism. In the whole-brain amyloid-negative subset, ACC amyloid was not correlated with age. As expected, the metabolism in an arbitrary region such as motor cortex that was not expected to decline with cognitive aging showed no correlation with age or ACC metabolism suggesting regional specificity. These findings motivate the search for the pathophysiology of aging-related ACC dysfunction to prevent, diagnose, and treat the decline in executive function associated with cognitive aging.

The Paintbrush Coverage Problem

Autonomous vehicles become more and more popular in our daily life. Mobile computing schemes to be installed on these vehicles have drawn a lot of recent research interest. In this paper, we address the important path-planning problem for autonomous vehicles. We introduce and formulate the novel <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">paintbrush coverage problem</i> . We present a theoretical study on the minimum trajectory length of a paintbrush to cover an arbitrary convex region, which is derived as a function of the area of the region and the size of the cover. Three commonly-used patrolling/scouting methods, namely boustrophedon, spiral, and sector, are manifested in details as the potential solutions to the paintbrush coverage problem. The theoretical minimum trajectory lengths any algorithm can achieve are also demonstrated as the benchmarks for different shapes of regions.

Optical properties and dust temperatures in clumpy diffuse medium

ABSTRACT In this study, we explore the impact of inhomogeneities in the spatial distribution of interstellar dust on spatial scales of ≤1 au caused by ion shadowing forces on the optical properties of diffuse interstellar medium (ISM) as well as on the dust temperature. We show that recently proposed possibility that interstellar dust grains in the diffuse ISM are grouped in spherical cloudlets (clumps) may significantly affect the observed optical properties of the diffuse ISM in comparison to that calculated under the commonly accepted assumption on the uniform dust/gas mixture if the size of clumps ≳0.1 au. We found that opacity of an arbitrary region of diffuse ISM quickly decreases with growth of dusty clumps. We also studied the dependence of opacity and dust temperature inside the dusty clumps on their size. We show that the clumps larger than 0.1 au are opaque for far-ultraviolet radiation. Dust temperature exhibits a gradient inside a clump, decreasing from the edge to the centre by several degrees for a clump of a size of 0.1 au and larger. We argue that dust temperatures and high opacity within clumps larger than 0.1 au may facilitate somewhat more efficient synthesis of molecules on surfaces of interstellar grains in the diffuse ISM than it was anticipated previously. On the other hand, the presence of clumps with sizes below 0.1 au makes small or negligible influence on the optical properties of the diffuse ISM in comparison to the case with uniform dust/gas mixture.

Finite-Time Neural Network Backstepping Control of an Uncertain Fractional-Order Duffing System With Input Saturation

In this paper, neural network (NN) control of fractional-order Duffing system (FODS) by suing backstepping method within finite-time in the presence of input saturation is investigated. A fractional-order filter with order lying on the interval (1,2) is used to estimate the virtual input together with its fractional derivative, and the estimation error tends to a small region in some finite time. Fractional-order law is designed for the parameter of NN, and an adaptive NN controller is given. The proposed method drives the tracking error tending to an arbitrary small region within finite time. The simulation results verify the validity of the proposed method.

The Riemann – Hilbert boundary value problem for elliptic systems of the orthogonal type in R&lt;sup&gt;3&lt;/sup&gt;

In this paper, a class of elliptic systems of four 1st order differential equations of the orthogonal type in R3 is considered. For such systems we study the issue of regularizability of the Riemann – Hilbert boundary value problem in an arbitrary limited simply-connected region with a smooth boundary in R3. Using the coefficients of the elliptic system and the matrix of the boundary operator, a special vector field is constructed, and its not entering the tangent plane in any point of the boundary provides the Lopatinski condition of the regularizability of the boundary value problem. The obtained condition permits to prove that the set of regularizable Riemann – Hilbert boundary value problems for the considered class of systems has two components of homotopic connectedness, and the index of an arbitrary regularizable problem equals to minus one.