Graphics Processing Research Articles

Machine learning classifiers for detection of glaucoma

Glaucoma is a disease that affects the optic nerve. This disease, over a period of time, can lead to loss of vision. Which is known as ‘silent thief of sight’. There are several methods in which the disease can be treated, if detected at an early stage It is not possible for any technology, including artificial intelligence, to replace a doctor. However, it is possible to develop a model based on several classical image processing algorithms, combined with artificial intelligence that can detect onset of glaucoma based on certain parameters of the retinal fundus. This model would play an important role in early detection of the disease and assist the doctor. The traditional methods to detect glaucoma, as efficient as they may be, are usually expensive, a machine learning approach to diagnose from fundus images and accurately classify its severity can be considered to be efficient. Here we propose support vector machine (SVM) method to segregate, train the models using a high-end graphics processor unit (GPU) and augment the hull convex approach to boost the accuracy of the image processing mechanisms along with distinguishing the different stages of glaucoma. A web application for the screening process has also been adopted.

UNRES-GPU for Physics-Based Coarse-Grained Simulations of Protein Systems at Biological Time- and Size-Scales.

The UNRES package for coarse-grained simulations, which has recently been optimized to treat large protein systems, has been implemented on Graphical Processor Units. An over 100-time speed up of the GPU code (run on an NVIDIA A100) with respect to the sequential code and an 8.5 speed-up with respect to the parallel (OpenMP) code (run on 32 cores of 2 AMD EPYC 7313 CPUs) has been achieved for large proteins (with size over 10,000 residues). Due to the averaging over the fine-grain degrees of freedom, 1 time unit of UNRES simulations is equivalent to about 1,000 time units of laboratory time, therefore millisecond time scale of large protein systems can be reached with the UNRES-GPU code. The complete source code of UNRES-GPU can be downloaded from https://projects.task.gda.pl/eurohpcpl-public/unres. Supplementary data are available at Bioinformatics online.

Streamline-based reactive transport modeling of uranium mining during in-situ leaching: Advantages and drawbacks

Reactive transport modeling is known to be computationally intensive when applied to 3D problems. Transforming sequential computing on the computer processor units (CPU) into parallelized computation on the high-performance parallel graphic processor units (GPU) is a classical approach to increasing computational performance. Another complementary approach is to decompose a complex 3D modeling problem into a set of simpler 1D problems using streamline approaches which can be easily parallelized, therefore reducing computation time. This paper investigates solutions to the equations governing dissolution and transport using streamlines coupled with a parallelization approach. In addition, an analytical solution to the dissolution and transfer equations of uranium describing the In-Situ Leaching (ISL) mining recovery is found using an approximation series to the 2nd order. The analytical solution is compared to the 1D numerical resolution along the streamlines and to the 3D simulation results superimposed on the streamline. Both approaches give similar results with a relative error of <2 % (2%). The proposed methodology is then applied to a case study in which the classical 3D resolution is compared to the newly suggested streamline solution, demonstrating that the streamline approach increases computational performances by a factor ranging from hundred to thousand depending on the complexity of the grid-block model.

Computer Simulation and Visualization of Wheel Tracks on Solid Surfaces in Virtual Environment

The paper describes methods and approaches for implementation of traces from virtual prototypes of wheeled technical vehicles on solid objects in three-dimensional virtual scenes. Use of particle systems hierarchically linked to wheel models is the basis of solutions developed. Original type of particle system is proposed, in which generation of elements is regulated by special control signal. Creation, processing, storage and destruction of particles is performed by means of the CUDA architecture with use of computing cores of modern graphics processor and video memory resources. Creating geometric model of each particle and its rendering is carried out on programmable graphics pipeline with own vertex, geometry and fragment shaders. Visualization of virtual environment with simulation of wheel tracks is performed in real time. Software modules were created based on developed methods and approaches. Approbation of them was carried out in virtual environment system VirSim and showed adequacy and effectiveness of proposed solutions for simulation of wheel tracks in virtual environment systems and training complexes.

PREPARATION AND PROPERTIES OF CERAMIC MATERIALS IN THE Ag6PS5I‒Ag7GeS5I SYSTEM

This work is devoted to the preparation of dense Ag+ conductive ceramic solid electrolytes in the Ag6PS5I‒Ag7GeS5I system and to the study of their electrical and optical properties. Using the non-destructive and simple method of optical diffuse reflectance spectroscopy, the value of the optical pseudogap energy of the Ag6+x(P1-xGex)S5I (x=0, 0.25, 0.5, 0.75, 1.0) solid solutions was obtained using the Kubelka-Munk function with the application of Tauc's graphic method. As a result of the performed mathematical analysis and graphic processing, it was established that the quaternary chalcohalogenides Ag6PS5I, Ag7GeS5I and solid solutions Ag6+x(P1-xGex)S5I (x=0.25, 0.5, 0.75) belong to direct-band semiconductors characterized by a forbidden optical transition . The study of electrical conductivity of Ag6+x(P1-xGex)S5I solid solutions (x=0, 0.25, 0.5, 0.75, 1.0) was carried out by the method of impedance spectroscopy, in the frequency 1×101–3×105 Hz and temperature 293–383 K ranges. For all ceramic materials prepared on the basis of Ag6+x(P1-xGex)S5I (x=0, 0.25, 0.5, 0.75, 1.0) solid solutions, an increase in the values of total electrical conductivity was observed, which is characteristic of materials with ionic electrical conductivity in the solid state.&#x0D; Keywords: argyrodites; solid solutions; ceramic; impedance spectroscopy; conductivity; optical properties; optical pseudogap.

Analysis Database Systems and Solve Medical Problems

Due to this mechanism, threshold synchronization is achieved with a minimum amount of time. It is usually designed for communication between scalar processors via shared memory. The main function of graphics processors since their inception has been graphics processing. Subsequently, after it became possible to program the processing of model vertices and pixels of rendered three-dimensional scenes using special programs (shaders), the architecture of graphic processors changed significantly. After the advent of the first general-purpose programmable graphics it became possible to process commands not only for graphic data in vector form, but also to perform ordinary calculations for arbitrary data on a variety of special cores, while implementing data parallelism.

Integration of Digital Technology in the Learning Process Through Problem-Based Learning Models

Digital technology is what most affects the education system today because of the aspects of effectiveness, efficiency, and attractiveness offered. The research approach used in this study was a quantitative research design. This study explores the integration of digital technology in students through quantitative data and enriches the data with qualitative analysis. The samples of this study were students of the education faculty majoring in elementary education at a university in Halu Oleo who were selected through convenience sampling. The first phase of the research is to conduct a problem-solving-based lecture process and then be given a standard scale and open-ended questions. Then the second phase, students are given a questionnaire given through a Google Form, which is sent along with all the information and an invitation to participate. The results of the study show the provision of digital technology-based learning content, especially in the problem-solving process, should consider quality, meaningful, and user-friendly materials. including spreadsheet applications, graphics processing; multimedia arrangements; mind/concept mapping; digital video processing; online class management, and online publishing; and the main reason for the inability to complete tasks is dominated by unsupported smart devices; inadequate programs or applications (software); inadequate internet connection; and the ability to operate the application is not supported.

PerFC: An Efficient 2D and 3D Perception Software-Hardware Framework for Mobile Cobot

In this work, we present an end-to-end software-hardware framework that supports both conventional hardware and software components and integrates machine learning object detectors without requiring an additional dedicated graphic processor unit (GPU). We design our framework to achieve real-time performance on the robot system, guarantee such performance on multiple computing devices, and concentrate on code reusability. We then utilize transfer learning strategies for 2D object detection and fuse them into depth images for 3D depth estimation. Lastly, we test the proposed framework on the Baxter robot with two 7-DOF arms and a four-wheel mobility base. The results show that the robot achieves real-time performance while executing other tasks (map building, localization, navigation, object detection, arm moving, and grasping) with available hardware like Intel onboard GPUs on distributed computers. Also, to comprehensively control, program, and monitor the robot system, we design and introduce an end-user application. The source code is available at https://github.com/tuantdang/perception_framework.

Visualizing research trends and thematic structure evolution of graphic design

The purpose of this research is to draw a scientific map and show the evolution of thematic structure of world’s graphic design. This is a scientometric study that uses quantitative and indicative methods of network analysis. To draw a scientific map of graphic design research literature, the data was extracted from the Web of Science database within a 2000–20 time period. The retrieved records were formatted using WC10.exe conversion software and then were mapped using Ravar Matrix, Ucinet, Netdraw, VOSviewer software. The findings revealed a growing trend in the number of articles published in graphic design and the most widely used subject areas in the field of graphic were identified. The field of graphic design was mapped based on the closeness centrality measure and its findings revealed that topics such as graphic sign, computer applications, molecular dynamics, computer architecture and graphics processing have the greatest impact on the map.

GVirtualXray (gVXR): Simulating X-ray radiographs and CT volumes of anthropomorphic phantoms

gVirtualXray (gVXR) is an open-source framework that relies on the Beer–Lambert law to simulate X-ray images in realtime on a graphics processor unit (GPU) using triangular meshes. We produced four Jupyter Notebooks to compare images simulated with gVXR and ground truth image of an anthropomorphic phantom: (i) an X-ray projection generated using a Monte Carlo simulation code, (ii) real digitally reconstructed radiographs (DRRs), (iii) computed tomography (CT) slices, and (iv) a real radiograph acquired with a clinical X-ray imaging system. Image registration was deployed in two Notebooks to align the simulated image on the corresponding ground truth image. We demonstrated that accurate images can be generated in milliseconds with gVirtualXray when scattering can be ignored.

Application of Computer Image Processing Technology in Visual Communication System

ABSTRACT This research paper aims to enhance the operation of visual communication systems by incorporating computer image processing technology. The study investigates the potential applications of visual communication systems by utilizing imaging equipment to capture distorted fringe images of height-modulated objects from different angles. Phase information closely associated with the object is then extracted from the images and utilized to obtain the object’s three-dimensional surface topography following calibration. The paper applies multi-frequency phase-shift profilometry and Fourier transforms profilometry to construct further a visual communication system based on computer graphics processing technology. The article contributes to visual communication systems by proposing a novel approach that incorporates computer image processing technology. This approach enhances the performance of visual communication systems, resulting in a more effective means of communication. The study’s experimental analysis demonstrates computer image processing technology’s superiority and ability to enhance the visual communication system’s application. The study’s key findings indicate that computer image processing technology can effectively enhance the effectiveness of visual communication systems. The incorporation of imaging equipment and phase extraction methods leads to the production of accurate three-dimensional surface topography data. Furthermore, multi-frequency phase-shift profilometry and Fourier transform profilometry aid in constructing a visual communication system based on computer graphics processing technology, which leads to an overall enhancement of the visual communication system’s application.

Prediction and Analysis of Heart Diseases Using Heterogeneous Computing Platform

The work in this paper helps study cardiac rhythms and the electrical activity of the heart for two of the most critical cardiac arrhythmias. Various consumer devices exist, but implementation of an appropriate device at a certain position on the body at a certain pressure point containing an enormous number of blood vessels and developing filtering techniques for the most accurate signal extraction from the heart is a challenging task. In this paper, we provide evidence of prediction and analysis of Atrial Fibrillation (AF) and Ventricular Fibrillation (VF). Long-term monitoring of diseases such as AF and VF occurrences is very important, as these will lead to occurrence of ischemic stroke, cardiac arrest and complete heart failure. The AF and VF signal classification accuracy are much higher when processed on a Graphics Processor Unit (GPU) than Central Processing Unit (CPU) or traditional Holter machines. The classifier COMMA-Z filter is applied to the highly-sensitive industry certified Bio PPG sensor placed at the earlobe and computed on GPU.

Simulation of X-ray projections on GPU: Benchmarking gVirtualXray with clinically realistic phantoms

Background and objectives: This study provides a quantitative comparison of images created using gVirtualXray (gVXR) to both Monte Carlo (MC) and real images of clinically realistic phantoms. gVirtualXray is an open-source framework that relies on the Beer-Lambert law to simulate X-ray images in realtime on a graphics processor unit (GPU) using triangular meshes.Methods: Images are generated with gVirtualXray and compared with a corresponding ground truth image of an anthropomorphic phantom: (i) an X-ray projection generated using a Monte Carlo simulation code, (ii) real digitally reconstructed radiographs (DRRs), (iii) computed tomography (CT) slices, and (iv) a real radiograph acquired with a clinical X-ray imaging system. When real images are involved, the simulations are used in an image registration framework so that the two images are aligned.Results: The mean absolute percentage error (MAPE) between the images simulated with gVirtualXray and MC is 3.12%, the zero-mean normalised cross-correlation (ZNCC) is 99.96% and the structural similarity index (SSIM) is 0.99. The run-time is 10 days for MC and 23 ms with gVirtualXray. Images simulated using surface models segmented from a CT scan of the Lungman chest phantom were similar to (i) DRRs computed from the CT volume and (ii) an actual digital radiograph. CT slices reconstructed from images simulated with gVirtualXray were comparable to the corresponding slices of the original CT volume.Conclusions: When scattering can be ignored, accurate images that would take days using MC can be generated in milliseconds with gVirtualXray. This speed of execution enables the use of repetitive simulations with varying parameters, e.g. to generate training data for a deep-learning algorithm, and to minimise the objective function of an optimisation problem in image registration. The use of surface models enables the combination of X-ray simulation with real-time soft-tissue deformation and character animation, which can be deployed in virtual reality applications.

AR/VR/3D Technologies: Enabling Future Business Growth and Scale-Up

Augmented Reality and Virtual Reality (AR/VR) technologies have been in existence for some time, but recent advancements in hardware, storage, graphics processing, and high-resolution displays have made them more accessible and practical. As a result, an increasing number of organizations are exploring these immersive technologies for various purposes. This research focuses on organizations that have already initiated their AR/VR/3D journey, either through experimentation or implementation, and aims to answer the following questions. ❖ How is AR/VR/3D being employed today? ❖ What are the latest trends in AR/VR/3D technology? ❖ How can firms either begin or evolve their initiatives by implementing AR/VR/3D tech?

Real-time musical performance across and within extended reality environments

The design of real-time digital musical instruments, both novel as well as the emulation of traditional musical performance practices using computer systems, has always required nearly imperceptible latencies between controlling gesture and expressed sound. Sound and music based experiences built within immersive visual rendered spaces often struggle to provide seamless articulate control of rendered dynamic output. Working across and within extended reality environments, the complexity of the challenge continues to evolve with current generations of hardware and software pushing core and graphics processors to their respective limits. The introduction of networked performance spaces has only added to the complexity, resulting in the need for additional latency mitigation strategies from both the technical and artistic domains. This paper will discuss recent creative and technical strategies and concerns in the design, development and use of real-time virtual musical instruments.

Computing the gravitational potential on nested meshes using the convolution method

Aims. Our aim is to derive a fast and accurate method for computing the gravitational potential of astrophysical objects with high contrasts in density, for which nested or adaptive meshes are required. Methods. We present an extension of the convolution method for computing the gravitational potential to the nested Cartesian grids. The method makes use of the convolution theorem to compute the gravitational potential using its integral form. Results. A comparison of our method with the iterative outside-in conjugate gradient and generalized minimal residual methods for solving the Poisson equation using nonspherically symmetric density configurations has shown a comparable performance in terms of the errors relative to the analytic solutions. However, the convolution method is characterized by several advantages and outperforms the considered iterative methods by factors 10–200 in terms of the runtime, especially when graphics processor units are utilized. The convolution method also shows an overall second-order convergence, except for the errors at the grid interfaces where the convergence is linear. Conclusions. High computational speed and ease in implementation can make the convolution method a preferred choice when using a large number of nested grids. The convolution method, however, becomes more computationally costly if the dipole moments of tightly spaced gravitating objects are to be considered at coarser grids.

APPROACH FOR CODE ANALYSIS TO ESTIMATE POWER CONSUMPTION OF CUDA CORE

The graphics processing unit is a popular computing device for achieving exascale performance in high-performance computing programs, which is used not only in graphics tasks, but also in computational tasks such as machine learning, scientific computing, and cryptography. With the help of a graphics processor, you can achieve significant speed and performance compared to the central processing unit. CUDA, Compute Unified Device Architecture, a graphics processing unit software development platform, allows developers to use the high-performance computing capabilities of graphics processing units to solve problems traditionally handled by central processing units. Even though the graphics processing unit has a relatively high power to performance ratio, it consumes a significant amount of power during computing. The paper proposes an approach for code analysis to estimate power consumption of CUDA core to improve the power efficiency of applications focused on computing on graphics processing units. The proposed approach makes it possible to estimate the power consumption of such applications without the need to run them on physical devices. The proposed approach is based on static analysis of the CUDA program and machine learning methods. To evaluate the effectiveness of the proposed approach, three graphics processing unit architectures were used: NVIDIA PASCAL, NVIDIA TURING, and NVIDIA AMPERE. The results of the experiments showed that for the NVIDIA AMPERE architecture, the proposed approach using decision trees makes it possible to achieve a determination coefficient of 0.9173. The results obtained confirm the effectiveness of the proposed code analysis method for estimating the power consumption of the CUDA core. This method can be useful for CUDA developers who want to improve the efficiency and power efficiency of their programs.

LSNet: Lightweight Spatial Boosting Network for Detecting Salient Objects in RGB-Thermal Images.

Most recent methods for RGB (red-green-blue)-thermal salient object detection (SOD) involve several floating-point operations and have numerous parameters, resulting in slow inference, especially on common processors, and impeding their deployment on mobile devices for practical applications. To address these problems, we propose a lightweight spatial boosting network (LSNet) for efficient RGB-thermal SOD with a lightweight MobileNetV2 backbone to replace a conventional backbone (e.g., VGG, ResNet). To improve feature extraction using a lightweight backbone, we propose a boundary boosting algorithm that optimizes the predicted saliency maps and reduces information collapse in low-dimensional features. The algorithm generates boundary maps based on predicted saliency maps without incurring additional calculations or complexity. As multimodality processing is essential for high-performance SOD, we adopt attentive feature distillation and selection and propose semantic and geometric transfer learning to enhance the backbone without increasing the complexity during testing. Experimental results demonstrate that the proposed LSNet achieves state-of-the-art performance compared with 14 RGB-thermal SOD methods on three datasets while improving the numbers of floating-point operations (1.025G) and parameters (5.39M), model size (22.1 MB), and inference speed (9.95 fps for PyTorch, batch size of 1, and Intel i5-7500 processor; 93.53 fps for PyTorch, batch size of 1, and NVIDIA TITAN V graphics processor; 936.68 fps for PyTorch, batch size of 20, and graphics processor; 538.01 fps for TensorRT and batch size of 1; and 903.01 fps for TensorRT/FP16 and batch size of 1). The code and results can be found from the link of https://github.com/zyrant/LSNet.

Direct Rendering of Three-Dimensional Objects Based on Perturbation Functions Using GPUs

The object of the study is a method of direct rendering of complex three-dimensional objects based on perturbation functions using graphics processors, using a variety of streaming multiprocessors. Direct rendering means that the visualization of functionally defined models takes place without their preliminary conversion to other formats, for example, into triangle grids. The research method is based on analytical geometry in space, differential geometry, interpolation theory and matrix theory, based on mathematical modeling and the theory of computing systems. The main conclusions of the study are: the possibility of direct rendering of functionally specified objects, when rendering it is important that the computing processors are not idle. The first problem that was solved was that different GPUs have different numbers of streaming multiprocessors. Therefore, it was necessary to choose during execution the optimal stage from which the work began. Thus, you can partially get rid of the problem with unused computing resources. The second problem, the balancing problem, was solved by using a large number of computing processors. For implementation, the CUDA parallel programming model was used, which, together with a set of software tools, allows implementing programs in the C language for execution on a GPU. The resulting system visualizes complex functionally defined objects with high resolution interactively. The dependence of performance on the computing power of graphics processors is investigated.

Обчислення сингулярного розкладу матриць з використанням графічного процесора

In this research paper we present an implementation of a singular value decomposition algorithm designed specifically for the graphics processing unit. It consists of two parts: orthogonal matrix decomposition and matrix diagonalization. Presented an implementation of bidiagonalization algorithm where we calculate the main bidiagonal matrix and two orthogonal multipliers using a series of House- holder transformations, as well as diagonalization algorithm with the help of Givens rotation matrices. Bothe these parts are implemented in jCUDA environment. Experiments have been conducted, the results of which have been thoroughly investigated on the matter of time consumption and calculations error. We’ve also compared our implementation with alternatives both on central and graphic processors.