Box Size Research Articles

Understanding the Temperature Dependence and Finite Size Effects in Ab Initio MD Simulations of the Hydrated Electron

The hydrated electron is of interest to both theorists and experimentalists as a paradigm solution-phase quantum system. Although the bulk of the theoretical work studying the hydrated electron is based on mixed quantum/classical (MQC) methods, recent advances in computer power have allowed several attempts to study this object using ab initio methods. The difficulty with employing ab initio methods for this system is that even with relatively inexpensive quantum chemistry methods such as density functional theory (DFT), such calculations are still limited to at most a few tens of water molecules and only a few picoseconds duration, leaving open the question as to whether the calculations are converged with respect to either system size or dynamical fluctuations. Moreover, the ab initio simulations of the hydrated electron that have been published to date have provided only limited analysis. Most works calculate the electron's vertical detachment energy, which can be compared to experiment, and occasionally the electronic absorption spectrum is also computed. Structural features, such as pair distribution functions, are rare in the literature, with the majority of the structural analysis being simple statements that the electron resides in a cavity, which are often based only on a small number of simulation snapshots. Importantly, there has been no ab initio work examining the temperature-dependent behavior of the hydrated electron, which has not been satisfactorily explained by MQC simulations. In this work, we attempt to remedy this situation by running DFT-based ab initio simulations of the hydrated electron as a function of both box size and temperature. We show that the calculated properties of the hydrated electron are not converged even with simulation sizes up to 128 water molecules and durations of several tens of picoseconds. The simulations show significant changes in the water coordination and solvation structure with box size. Our temperature-dependent simulations predict a red-shift of the absorption spectrum (computed using TD-DFT with an optimally tuned range-separated hybrid functional) with increasing temperature, but the magnitude of the predicted red-shift is larger than that observed experimentally, and the absolute position of the calculated spectra are off by over half an eV. The spectral red-shift at high temperatures is accompanied by both a partial loss of structure of the electron's central cavity and an increased radius of gyration that pushes electron density onto and beyond the first solvation shell. Overall, although ab initio simulations can provide some insights into the temperature-dependent behavior of the hydrated electron, the simulation sizes and level of quantum chemistry theory that are currently accessible are inadequate for correctly describing the experimental properties of this fascinating object.

Accelerated 2D Classification With ISAC Using GPUs.

A widely used approach to analyze single particles in electron microscopy data is 2D classification. This process is very computationally expensive, especially when large data sets are analyzed. In this paper we present GPU ISAC, a newly developed, GPU-accelerated version of the established Iterative Stable Alignment and Clustering (ISAC) algorithm for 2D images and generating class averages. While the previously existing implementation of ISAC relied on a computer cluster, GPU ISAC enables users to produce high quality 2D class averages from large-scale data sets on a single desktop machine equipped with affordable, consumer-grade GPUs such as Nvidia GeForce GTX 1080 TI cards. With only two such cards GPU ISAC matches the performance of twelve high end cluster nodes and, using high performance GPUs, is able to produce class averages from a million particles in between six to thirteen hours, depending on data set quality and box size. We also show GPU ISAC to scale linearly in all input dimensions, and thereby capable of scaling well with the increasing data load demand of future data sets. Further user experience improvements integrate GPU ISAC seamlessly into the existing SPHIRE GUI, as well as the TranSPHIRE on-the-fly processing pipeline. It is open source and can be downloaded at https://gitlab.gwdg.de/mpi-dortmund/sphire/cuISAC/

The pseudoscalar meson and baryon octet interaction with strangeness S = −2 in the unitary coupled-channel approximation

The interaction of the pseudoscalar meson and the baryon octet with strangeness and isospin is investigated by solving the Bethe–Salpeter equation in the infinite and finite volume respectively. It is found that there is a resonance state generated dynamically, which owns a mass of about 1550 MeV and a large decay width of 120–200 MeV. This resonance state couples strongly to the π channel. Therefore, it might not correspond to the particle announced by Belle collaboration. At the same time, this problem is studied in the finite volume, and an energy level at 1570 MeV is obtained, which is between the and thresholds and independent of the cubic box size.

Spurious poles in a finite volume

Using effective-range expansion for the two-body amplitudes may generate spurious sub-threshold poles outside of the convergence range of the expansion. In the infinite volume, the emergence of such poles leads to the inconsistencies in the three-body equations, e.g., to the breakdown of unitarity. We investigate the effect of the spurious poles on the three-body quantization condition in a finite volume and show that it leads to a peculiar dependence of the energy levels on the box size L. Furthermore, within a simple model, it is demonstrated that the procedure for the removal of these poles, which was recently proposed in ref. [1] in the infinite volume, can be adapted to the finite-volume calculations. The structure of the exact energy levels is reproduced with an accuracy that systematically improves order by order in the EFT expansion.

SIMULASI PENDUGAAN LONGSOR: PENGARUH INTENSITAS HUJAN PADA TANAH DENGAN TEKSTUR DAN KANDUNGAN BAHAN ORGANIK YANG BERBEDA

Natural factors and human actions can cause landslides. Natural factors include high rainfall intensity and mountainous areas or hills with steep slopes. Human actions that can cause landslides are land conversion and inappropriate land management. This study aimed to explore the effect of rain intensity on soils with different textures and organic matter content. The landslide simulation study was conducted at the Agro Techno Park and Experimental Field, Faculty of Agriculture, Brawijaya University. This study used an acrylic box (size 100 cm x 50 cm x 75 cm) and aluminium plate as a base, which was adjusted to a slope of 50. There were eight treatment combinations consisting of three factors: soil texture (silty and clayey), rainfall intensity (70 mm hour-1 and 120 mm hour-1 ), and organic matter application (without and with organic matter application). The variables measured were landslide duration and volume. The results showed that the duration and volume of landslide materials are largerly influenced by rainfall intensity. Adding organic matter to the soil at any rainfall intensity may decrease landslide duration and volume by half compared to without organic matter application. Among the three factors examined (soil texture, organic matter application and rainfall intensity), rainfall intensity seems to be the most determining factor in landslide duration and volume.

On the existence of global‐in‐time weak solutions and scaling laws for Kolmogorov's two‐equation model for turbulence

AbstractThis paper is concerned with Kolmogorov's two‐equation model for turbulence in involving the mean velocity u, the pressure p, an average frequency , and a mean turbulent kinetic energy k. We consider the system with space‐periodic boundary conditions in a cube , which is a good choice for studying the decay of free turbulent motion sufficiently far away from boundaries. In particular, this choice is compatible with the rich set of similarity transformations for turbulence. The main part of this work consists in proving existence of global weak solutions of this model. For this we approximate the system by adding a suitable regularizing r‐Laplacian and invoke existence result for evolutionary equations with pseudo‐monotone operators. An important point constitutes the derivation of pointwise a priori estimates for ω (upper and lower) and k (only lower) that are independent of the box size a, thus allow us to control the parabolicity of the diffusion operators.

RapidLayout: Fast Hard Block Placement of FPGA-optimized Systolic Arrays Using Evolutionary Algorithm

Evolutionary algorithms can outperform conventional placement algorithms such as simulated annealing, analytical placement, and manual placement on runtime, wirelength, pipelining cost, and clock frequency when mapping hard block intensive designs such as systolic arrays on Xilinx UltraScale+ FPGAs. For certain hard-block intensive designs, the commercial-grade Xilinx Vivado CAD tool cannot provide legal routing solutions without tedious manual placement constraints. Instead, we formulate hard block placement as a multi-objective optimization problem that targets wirelength squared and bounding box size. We build an end-to-end placement-and-routing flow called RapidLayout using the Xilinx RapidWright framework. RapidLayout runs 5–6 \( \times \) faster than Vivado with manual constraints and eliminates the weeks-long effort to manually generate placement constraints. RapidLayout enables transfer learning from similar devices and bootstrapping from much smaller devices. Transfer learning in the UltraScale+ family achieves 11–14 \( \times \) shorter runtime and bootstrapping from a 97% smaller device delivers 2.1–3.2 \( \times \) faster optimizations. RapidLayout outperforms (1) a tuned simulated annealer by 2.7–30.8 \( \times \) in runtime while achieving similar quality of results, (2) VPR by 1.5 \( \times \) in runtime, 1.9–2.4 \( \times \) in wirelength, and 3–4 \( \times \) in bounding box size, while also (3) beating the analytical placer UTPlaceF by 9.3 \( \times \) in runtime, 1.8–2.2 \( \times \) in wirelength, and 2–2.7 \( \times \) in bounding box size.

AUTOMATIC DETECTION AND GRADING OF DIABETIC MACULAR EDEMA BASED ON A DEEP NEURAL NETWORK.

To solve the problem of automatic grading of macular edema in retinal images in a more stable and reliable way and reduce the workload of ophthalmologists, an automatic detection and grading method of diabetic macular edema based on a deep neural network is proposed. The enhanced green channels of fundus images are input into the YOLO network for training and testing. Diabetic macular edema is graded according to the distance of the macula and hard exudate. We used multiscale feature fusion to form more comprehensive features on different grain images to improve the effect of hard exudate detection. We adopted K-means++ algorithm to cluster anchor box size and use loss of the original network to guide the regression of hard exudate bounding box and improve the regression accuracy of anchor boxes. We increased the diversity of samples for sample training by data augmentation, including cropping, flipping, and rotating of fundus images, so that each batch of training data can better represent the distribution of samples. The detection accuracy of the proposed method can reach 96% on the MESSIDOR data set. The detection rates of hard exudate with high, median, and low probability are 100%, 79.12%, and 60.40%, respectively. The proposed method exhibits a very good detection stability on healthy and diseased fundus images.

Traffic Sign Board Detection Using Single Shot Detection (SSD)

Abstract: Traffic sign board detection (TSBD) is a significant portion of intelligent transportation system (ITS). Being able to identify traffic signals more accurately and effectively can improve safe driving .Due to increase in technology there are autonomous vehicles . The traffic sign recognition process includes two parts: detection and classification. In this paper, we use an object detection algorithm called SSD to detect the traffic signs. This convolutional neural network uses multiple feature maps to detect objects. For the traffic sign is very small to the whole picture, the SSD model has been improved to have a better detection result of traffic signs. In the experiments, the model has been simplified and the size of the prior box has been modified. The improved network has a good detection effect on small targets. The results on the test data set show that the proposed algorithm performs well for single-target, multi-target and dark-light images. The precision and recall on the test data set are 91.09%, and 88.06%. Keywords: Automatic traffic sign board Detection, SSD, Image processing, Text alert.

The role of surface chemistry on CO2 adsorption in biomass-derived porous carbons by experimental results and molecular dynamics simulations

Biomass-derived porous carbons have been considered one of the most effective adsorbents for CO2 capture, due to their porous structure and high specific surface area. In this study, we successfully synthesized porous carbon from celery biomass and examined the effect of external adsorption parameters including time, temperature, and pressure on CO2 uptake in experimental and molecular dynamics (MD) simulations. Furthermore, the influence of carbon’s surface chemistry (carboxyl and hydroxyl functionalities) and nitrogen type on CO2 capture were investigated utilizing MD simulations. The results showed that pyridinic nitrogen has a greater tendency to adsorb CO2 than graphitic. It was found that the simultaneous presence of these two types of nitrogen has a greater effect on the CO2 sorption than the individual presence of each in the structure. It was also revealed that the addition of carboxyl groups (O=C–OH) to the carbon matrix enhances CO2 capture by about 10%. Additionally, by increasing the simulation time and the size of the simulation box, the average absolute relative error for simulation results of optimal structure declined to 16%, which is an acceptable value and makes the simulation process reliable to predict adsorption capacity under various conditions.

AI framework with computational box counting and Integer programming removes quantization error in fractal dimension analysis of optical images

Because many Chemical Engineering phenomena and processes involve distinctive shapes and structures, fractal dimension (FD) analysis is of great interest to the Chemical Engineering community due to its utility in providing a statistical index of shape complexity, which not only quantifies image structure but also helps explain functional properties. Past box counting (BC) methods for estimation of FDs are inconsistent due to quantization error (QE) introduced from image rotation and translation. In this work, we propose a systematic and automatic artificial intelligence (AI) framework that consistently estimates FDs of relevance to different fields of Chemical Engineering without QE by integrating image preprocessing, set-covering optimization-based BC, and regression analysis. As a result of the deterministic optimization technique, FD estimations remain consistent for all images regardless of image rotation and translation. The results of image datasets obtained from fields such as interfacial science, biomedical engineering, human anatomy, among others, demonstrate efficient box size specification to maximize regression effectiveness and FD calculation accuracy for a variety of images. Overall, the proposed AI framework offers a new means of estimating FD accurately and efficiently for optical images of interest to the Chemical Engineering community.

Improving the Algorithm Study of YOLO in Steel Surface Defect Detection

To solve the problem of low detection accuracy caused by background interference and diverse target forms, a series of improvements are proposed to improve the detection accuracy. According to the various characteristics of steel surface defects, this paper presents the K-Means clustering algorithm to optimize the clustering results and quickly and accurately obtain the size of the prior box. In view of the small proportion of the target defect area in the overall image and background interference, a two-way attention module (TWA-Block) is proposed to establish the long-distance dependence of the spatial domain and channel domain features, and a background suppression function is designed to realize the division of defect areas. Experiments of the proposed improvements in the NEU-DET dataset based on the YOLO series model show that the detection accuracy of all the improved YOLO series models has improved, and the number of parameters will not increase substantially

Air microfluidics-enabled soft robotic transtibial prosthesis socket liner toward dynamic management of residual limb contact pressure and volume fluctuation.

Residual limb volume fluctuation and the resulting contact pressures are some of the key factors leading to skin ulcerations, suboptimal prosthetic functioning, pain, and diminishing quality of life of transtibial amputees. Self-management of socket fit is complicated by peripheral neuropathy, reducing the perception of pressure and pain in the residual limb. We introduce a novel proof-of-concept for a transtibial prosthetic socket liner with the potential to dynamically adjust the fit between the limb and socket. The core of the technology is a small air microfluidic chip (10 cm3 and 10 g) with 10 on-chip valves that enable sequential pressurizing of 10 actuators in custom sizes to match the pressures required by the residual limb's unique anatomy. The microfluidic chip largely reduced the number of electromechanical solenoid valves needed for sequential control of 10 actuators (2 instead of 10 valves), resulting in the reduction of the required power, size, mass, and cost of the control box toward an affordable and wearable prosthetic socket. Proof-of-concept testing demonstrated that the applied pressures can be varied in the desired sequence and to redistribute pressure. Future work will focus on integrating the system with biofidelic prosthetic sockets and residual limb models to investigate the ability to redistribute pressure away from pressure-sensitive regions (e.g., fibular head) to pressure tolerant areas. Overall, the dynamic prosthesis socket liner is very encouraging for creating a dynamic socket fit system that can be seamlessly integrated with existing socket fabrication methods for managing residual limb volume fluctuations and contact pressure.

2차원 영상에서의 깊이 추정을 위한 객체 인식 제어 방법

In this paper, we propose a method for controlling a manipulator using a single 2D camera and an experimental system. Typically, to control the manipulator, the camera is constantly moved using a motor to navigate or detect the reference object in a fixed frame. The proposed control method does not require conventional high-performance depth cameras and can derive depth values from images acquired using a low-cost 2D camera. A real-time image frame detects objects that are the criteria and configures the learning model and communication node to draw an object-bounding box accordingly. In this work, we convert the change in size of this bounding box into a depth value. When the depth value is thus derived from the 2D image, the target position value of the manipulator can be determined. Finally, the manipulator system is configured with detection control.

Using lightweight convolutional neural network to track vibration displacement in rotating body video

The non-contact visual displacement measurement methods can detect problems such as rigid body deformation and structural wear of structural bodies due to long-term service life. However, the traditional visual vibration measurement methods are limited by the inherent sampling constraint of ordinary image sensors, which cannot achieve efficient vibration target identification and correlation tracking for rotating body targets in low-resolution videos under the premise of high sampling rate. Therefore, this paper uses high-speed industrial camera as the acquisition medium, introduces deep convolutional neural networks into the field of visual vibration measurement, on this basis, weighs the target detection accuracy and displacement tracking speed, uses the designed lightweight convolutional neural networks model to vibration displacement tracking in rotating body videos. We ensure that the computational efficiency and parameter quantity of the convolutional networks are solved with low loss accuracy. First of all, we take the lightweight convolutional neural network as the backbone network, replace the standard convolutional neural network with depthwise separable convolution and pointwise convolution. Considering the speed and accuracy advantages of deep learning algorithms for video object tracking, we estimate the heat map, object center offsets and bounding box sizes use an anchor-free detection algorithm on the network framework. In order to prevent the loss of vibrating target identity in rotating body videos, we use re-identification(re-ID) method to strengthen the correlation of target displacement between adjacent frames. In experiments with traditional visual vibration measurement and recent deep learning measurement methods for performance testing, the network model we designed can demonstrate absolute advantages that traditional convolutional neural networks do not have. On the one hand, our comparison of time–frequency characteristics at different speeds shows that the vibration displacement curve regressed by the lightweight convolutional neural network has a high degree of fit with the displacement signal obtained by the eddy current sensor;on the other hand, when the target object is blurred, the generalization ability of our algorithm is proved, which also reflects the engineering application value of visual vibration measurement in the field of vibration displacement tracking of rotating bodies.

Cross-talk elimination for lenslet array near eye display based on eye-gaze tracking.

Lenslet array (LA) near-eye displays (NEDs) are a recent technical development that creates a virtual image in the field of view of one or both eyes. A problem occurs when the user's pupil moves out of the LA-NED eye box (i.e., cross-talk) making the image look doubled or ghosted. It negatively impacts the user experience. Although eye-gaze tracking can mitigate this problem, the effect of the solution has not been studied to understand the impact of pupil size and human perception. In this paper, we redefine the cross-talk region as the practical pupil movable region (PPMR50), which differs from eye box size because it considers pupil size and human visual perception. To evaluate the effect of eye-gaze tracking on subjective image quality, three user studies were conducted. From the results, PPMR50 was found to be consistent with human perception, and cross-talk elimination via eye-gaze tracking was better understood in a static gaze scenario. Although the system latency prevented the complete elimination of cross-talk for fast movements or large pupil changes, the problem was greatly alleviated. We also analyzed system delays based on PPMR50, which we newly defined in this paper and provided an optimization scheme to meet the maximum eyeball rotation speed.

Unveiling the Molecular Origin of Vapor-Liquid Phase Transition of Bulk and Confined Fluids

At temperatures below the critical temperature, discontinuities in the isotherms are one critical issue in the design and construction of separation units, affecting the level of confidence for a prediction of vapor–liquid equilibriums and phase transitions. In this work, we study the molecular mechanisms of fluids that involve the vapor–liquid phase transition in bulk and confinement, utilizing grand canonical (GCE) and meso-canonical (MCE) ensembles of the Monte Carlo simulation. Different geometries of the mesopores, including slit, cylindrical, and spherical, were studied. During phase transitions, condensation/evaporation hysteretic isotherms can be detected by GCE simulation, whereas employing MCE simulation allows us to investigate van der Waals (vdW) loop with a vapor spinodal point, intermediate states, and a liquid spinodal point in the isotherms. Depending on the system, the size of the simulation box, and the MCE method, we are able to identify three distinct groups of vdW-type isotherms for the first time: (1) a smooth S-shaped loop, (2) a stepwise S-shaped loop, and (3) a stepwise S-shaped loop with just a vertical segment. The first isotherm type is noticed in the bulk and pores having small box sizes, in which vapor and liquid phases are close and not clearly identified. The second and the third types occurred in the bulk, cylindrical, and slit mesopores with sufficiently large spaces, where vapor and liquid phases are distinctly separated. Results from our studies provide an insight analysis into vapor–liquid phase transitions, elucidating the effect of the confinement of fluid behaviors in a visual manner.

Numerical study of cosmic ray confinement through dust resonant drag instabilities

ABSTRACT We investigate the possibility of cosmic ray (CR) confinement by charged dust grains through resonant drag instabilities (RDIs). We perform magnetohydrodynamic particle-in-cell simulations of magnetized gas mixed with charged dust and cosmic rays, with the gyro-radii of dust and GeV CRs on ∼au scales fully resolved. As a first study, we focus on one type of RDI wherein charged grains drift super-Alfvénically, with Lorentz forces strongly dominating over drag forces. Dust grains are unstable to the RDIs and form concentrated columns and sheets, whose scale grows until saturating at the simulation box size. Initially perfectly streaming CRs are strongly scattered by RDI-excited Alfvén waves, with the growth rate of the CR perpendicular velocity components equaling the growth rate of magnetic field perturbations. These rates are well-predicted by analytic linear theory. CRs finally become isotropized and drift at least at ∼vA by unidirectional Alfvén waves excited by the RDIs, with a uniform distribution of the pitch angle cosine μ and a flat profile of the CR pitch angle diffusion coefficient Dμμ around μ = 0, without the ‘90○ pitch angle problem.’ With CR feedback on the gas included, Dμμ decreases by a factor of a few, indicating a lower CR scattering rate, because the backreaction on the RDI from the CR pressure adds extra wave damping, leading to lower quasi-steady-state scattering rates. Our study demonstrates that the dust-induced CR confinement can be very important under certain conditions, e.g. the dusty circumgalactic medium around quasars or superluminous galaxies.

Vehicle detection method for satellite videos based on enhanced vehicle features

Vehicle detection by satellite videos is important for urban intelligent traffic construction. A satellite videos vehicle detection method is proposed to solve the problems of the surrounding background and the vehicles having a similar color, and the low number of pixels occupied by vehicles. First, we design a weighted normalization sparse autoencoder network, highlighting the effective features and strengthening the expression of the vehicle features by introducing weighted normalization sparse autoencoders into the backbone. Second, we design a method for the early fusion of spatial information transmission, realizing the enhancement of spatial information in the deep feature map through atrous convolution. Finally, we propose K-means++ based on adjusted cosine similarity, stretching the anchor box size by introducing the adjusted cosine similarity calculation. The experimental results show that the average accuracy of this method in satellite videos vehicle detection reached 75.7%, and the recall rate reached 89.6%. This method can effectively detect vehicles in satellite videos, and the effect is significantly improved compared with other methods.

UAV Imagery for Automatic Multi-Element Recognition and Detection of Road Traffic Elements

Road traffic elements comprise an important part of roads and represent the main content involved in the construction of a basic traffic geographic information database, which is particularly important for the development of basic traffic geographic information. However, the following problems still exist for the extraction of traffic elements: insufficient data, complex scenarios, small targets, and incomplete element information. Therefore, a set of road traffic multielement remote sensing image datasets obtained by unmanned aerial vehicles (UAVs) is produced, and an improved YOLOv4 network algorithm combined with an attention mechanism is proposed to automatically recognize and detect multiple elements of road traffic in UAV imagery. First, the scale range of different objects in the datasets is counted, and then the size of the candidate box is obtained by the k-means clustering method. Second, mosaic data augmentation technology is used to increase the number of trained road traffic multielement datasets. Then, by integrating the efficient channel attention (ECA) mechanism into the two effective feature layers extracted from the YOLOv4 backbone network and the upsampling results, the network focuses on the feature information and then trains the datasets. At the same time, the complete intersection over union (CIoU) loss function is used to consider the geometric relationship between the object and the test object, to solve the overlapping problem of the juxtaposed dense test element anchor boxes, and to reduce the rate of missed detection. Finally, the mean average precision (mAP) is calculated to evaluate the experimental effect. The experimental results show that the mAP value of the proposed method is 90.45%, which is 15.80% better than the average accuracy of the original YOLOv4 network. The average detection accuracy of zebra crossings, bus stations, and roadside parking spaces is improved by 12.52%, 22.82%, and 12.09%, respectively. The comparison experiments and ablation experiments proved that the proposed method can realize the automatic recognition and detection of multiple elements of road traffic, and provide a new solution for constructing a basic traffic geographic information database.