Search
Filters
total: 79
Search results for: CONVOLUTION
-
Longitudinal drug synergy assessment using convolutional neural network image-decoding of glioblastoma single-spheroid cultures
PublicationAbstract Background In recent years, drug combinations have become increasingly popular to improve therapeutic outcomes in various diseases, including difficult to cure cancers such as the brain cancer glioblastoma. Assessing the interaction between drugs over time is critical for predicting drug combination effectiveness and minimizing the risk of therapy resistance. However, as viability readouts of drug combination experiments...
-
Flexible Knowledge–Vision–Integration Platform for Personal Protective Equipment Detection and Classification Using Hierarchical Convolutional Neural Networks and Active Leaning
PublicationThis work is part of an effort to develop of a Knowledge-Vision Integration Platform for Hazard Control (KVIP-HC) in industrial workplaces, adaptable to a wide range of industrial environments. The paper focuses on hazards resulted from the non-use of personal protective equipment (PPE). The objective is to test the capability of the platform to adapt to different industrial environments by simulating the process of randomly selecting...
-
Activation maps of convolutional neural networks as a tool for brain degeneration tracking in early diagnosis of dementia in Parkinson's disease based on magnetic resonance imaging
Publication -
Colvolutional calibration of AFM probe
Open Research DataAtomic force microscopy is based on the interaction of the examined surface with a probe of a pyramidal shape, tipped with a sharp end with a radius of curvature ranging from single nanometers to hundreds of nanometers. The resolution of the obtained image is of course dependent on the above-mentioned geometric size, and the resulting image is a convolutional...
-
Image Processing Techniques for Distributed Grid Applications
PublicationParallel approaches to 2D and 3D convolution processing of series of images have been presented. A distributed, practically oriented, 2D spatial convolution scheme has been elaborated and extended into the temporal domain. Complexity of the scheme has been determined and analysed with respect to coefficients in convolution kernels. Possibilities of parallelisation of the convolution operations have been analysed and the results...
-
Inverse Flood Routing Using Simplified Flow Equations
PublicationThe paper considers the problem of inverse flood routing in reservoir operation strategy. The aim of the work is to investigate the possibility of determining the hydrograph at the upstream end based on the hydrograph required at the downstream end using simplified open channel flow models. To accomplish this, the linear kinematic wave equation, the diffusive wave equation and the linear Muskingum equation are considered. To achieve...
-
Discrete Green's function approach to disjoint domain simulations in 3D FDTD method
PublicationA discrete Green’s function (DGF) approach to couple 3D FDTD subdomains is developed. The total-field/scattered-field subdomains are simulated using the explicit FDTD method whilst interaction between them is computed as a convolution of the DGF with equivalent current sources measured over Huygens surfaces. In the developed method, the DGF waveforms are truncated using the Hann’s window. The error varies in the range -65 to -40...
-
Characteristics of an image sensor with early-vision processing fabricated in standard 0.35 µm CMOS technology
PublicationThe article presents measurement results of prototype integrated circuits for acquisition and processing of images in real time. In order to verify a new concept of circuit solutions of analogue image processors, experimental integrated circuits were fabricated. The integrated circuits, designed in a standard 0.35 µm CMOS technology, contain the image sensor and analogue processors that perform low-level convolution-based image...
-
Acceleration of the DGF-FDTD method on GPU using the CUDA technology
PublicationWe present a parallel implementation of the discrete Green's function formulation of the finite-difference time-domain (DGF-FDTD) method on a graphics processing unit (GPU). The compute unified device architecture (CUDA) parallel computing platform is applied in the developed implementation. For the sake of example, arrays of Yagi-Uda antennas were simulated with the use of DGF-FDTD on GPU. The efficiency of parallel computations...
-
Analytical Expression for the Time-Domain Discrete Green's Function of a Plane Wave Propagating in the 2-D FDTD Grid
PublicationIn this letter, a new closed-form expression for the time-domain discrete Green's function (DGF) of a plane wave propagating in the 2-D finite-difference time-domain (FDTD) grid is derived. For the sake of its verification, the time-domain implementation of the analytic field propagator (AFP) technique was developed for the plane wave injection in 2-D total-field/scattered-field (TFSF) FDTD simulations. Such an implementation of...
-
Analysis of radiation and scattering problems with the use of hybrid techniques based on the discrete Green's function formulation of the FDTD method
PublicationIn this contribution, simulation scenarios are presented which take advantage of the hybrid techniques based on the discrete Green's function formulation of the finite-difference time-domain (DGF-FDTD) method. DGF-FDTD solutions are compatible with the finite-difference grid and can be applied for perfect hybridization of the FDTD method. The following techniques are considered: (i) DGF-FDTD for antenna simulations, (ii) DGF-based...
-
Residual MobileNets
PublicationAs modern convolutional neural networks become increasingly deeper, they also become slower and require high computational resources beyond the capabilities of many mobile and embedded platforms. To address this challenge, much of the recent research has focused on reducing the model size and computational complexity. In this paper, we propose a novel residual depth-separable convolution block, which is an improvement of the basic...
-
Analytical Expression for the Time-Domain Green's Function of a Discrete Plane Wave Propagating in the 3-D FDTD Grid
PublicationIn this paper, a closed-form expression for the time-domain dyadic Green’s function of a discrete plane wave (DPW) propagating in a 3-D finite-difference time-domain (FDTD) grid is derived. In order to verify our findings, the time-domain implementation of the DPW-injection technique is developed with the use of the derived expression for 3-D total-field/scattered-field (TFSF) FDTD simulations. This implementation requires computations...
-
The estimation of fish lenght distribution from its acoustical measures
PublicationThe paper concerns the problem of estimating fish length PDF from its target strength PDF obtained from acoustic surveys. It has been shown that the target strength of single fish can be treated in the first approximation as a finction of two variables: one, which depends on fish size and the other, which depends on its angular orientation (aspect). Extending this simplified relationship to the case of fish populations allows to...
-
Application of the discrete Green's function-based antenna simulations for excitation of the total-field/scattered-field interface in the FDTD method
PublicationIn this article, the discrete Green's function formulation of the finite-difference time-domain (DGF-FDTD) method is proposed for simulation of wire antennas irradiating inhomogeneous dielectric scatterers. Surface equivalence theorem in the discrete domain is used to separate the problem into an inhomogeneous domain and a wire antenna that are simulated with the use of FDTD and DGF-FDTD, respectively. Then, the excitation of the...
-
Simulation of Wave Propagation in Media Described by Fractional-Order Models
PublicationIn this paper, algorithms for simulation of the wave propagation in electromagnetic media described by fractional-order (FO) models (FOMs) are presented. Initially, fractional calculus and FO Maxwell's equations are introduced. The problem of the wave propagation is formulated for media described by FOMs. Then, algorithms for simulation of the non-monochromatic wave propagation are presented which employ computations in the time...
-
Performance assessment of OpenMP constructs and benchmarks using modern compilers and multi-core CPUs
PublicationConsidering ongoing developments of both modern CPUs, especially in the context of increasing numbers of cores, cache memory and architectures as well as compilers there is a constant need for benchmarking representative and frequently run workloads. The key metric is speed-up as the computational power of modern CPUs stems mainly from using multiple cores. In this paper, we show and discuss results from running codes such as:...
-
Dobór głównych parametrów instalacji grzewczych spalinowych siłowni okrętowych z wykorzystaniem metod i modeli probabilistycznych
PublicationPraca doktorska przedstawia nowatorską metodę projektowania okrętowych instalacji grzewczych. Metoda ta powstała dzięki analizie statystycznej danych zebranych z kilkunastu statków dotyczących między innymi wytwarzania pary w kotłach utylizacyjnych i zapotrzebowania na ciepło przez odbiorniki pary. Zaproponowany w pracy proces projektowy wykorzystuje równania korelacji wynikające z analizy statystycznej. Wytwarzanie ciepła oraz...
-
FDTD Simulations on Disjoint Domains with the Use of Discrete Green's Function Diakoptics
PublicationA discrete Green's function (DGF) approach to couple disjoint domains in the finite-difference time-domain (FDTD) grid is developed. In this method, total-field/scattered-field (TFSF) FDTD domains are associated with simulated objects whereas the interaction between them is modeled with the use of the DGF propagator. Hence, source and scatterer are simulated in separate domains and updating of vacuum cells, being of little interest,...
-
Parallel Implementation of the Discrete Green's Function Formulation of the FDTD Method on a Multicore Central Processing Unit
PublicationParallel implementation of the discrete Green's function formulation of the finite-difference time-domain (DGF-FDTD) method was developed on a multicore central processing unit. DGF-FDTD avoids computations of the electromagnetic field in free-space cells and does not require domain termination by absorbing boundary conditions. Computed DGF-FDTD solutions are compatible with the FDTD grid enabling the perfect hybridization of FDTD...
-
Geometric analogue of holographic reduced representation
PublicationHolographic reduced representations (HRRs) are distributed representations of cognitive structuresbased on superpositions of convolution-bound n-tuples. Restricting HRRs to n-tuples consisting of 1,one reinterprets the variable binding as a representation of the additive group of binary n-tupleswith addition modulo 2. Since convolutions are not defined for vectors, the HRRs cannot be directlyassociated with geometric structures....
-
An inverse algorithm for contact heat conduction problems with an interfacial heat source based on a first-order thermocouple model
PublicationInverse problems of contact heat conduction with an interfacial heat source are common in various fields of science, engineering and technology. In this study, an algorithm for their solution is developed based on an inverse parametric optimisation method with an impulse response function describing the heat partition and contact heat transfer. A first-order thermocouple model with a time constant parameter is embedded in the impulse...
-
Applications of the discrete green's function in the finite-difference time-domain method
PublicationIn this paper, applications of the discrete Green's function (DGF) in the three-dimensional (3-D) finite-difference time-domain (FDTD) method are presented. The FDTD method on disjoint domains was developed employing DGF to couple the subdomains as well as to compute the electromagnetic field outside these subdomains. Hence, source and scatterer are simulated in separate subdomains and updating of vacuum cells, being of little...
-
Acceleration of the Discrete Green’s Function Formulation of the FDTD Method Based on Recurrence Schemes
PublicationIn this paper, we investigate an acceleration of the discrete Green's function (DGF) formulation of the FDTD method (DGF-FDTD) with the use of recurrence schemes. The DGF-FDTD method allows one to compute FDTD solutions as a convolution of the excitation with the DGF kernel. Hence, it does not require to execute a leapfrog time-stepping scheme in a whole computational domain for this purpose. Until recently, the DGF generation...
-
Deep Learning Optimization for Edge Devices: Analysis of Training Quantization Parameters
PublicationThis paper focuses on convolution neural network quantization problem. The quantization has a distinct stage of data conversion from floating-point into integer-point numbers. In general, the process of quantization is associated with the reduction of the matrix dimension via limited precision of the numbers. However, the training and inference stages of deep learning neural network are limited by the space of the memory and a...
-
Digital structures for high-speed signal processing
PublicationThe work covers several issues of realization of digital structures for pipelined processing of real and complex signals with the use of binary arithmetic and residue arithmetic. Basic rules of performing operations in residue arithmetic are presented along with selected residue number systems for processing of complex signals and computation of convolution. Subsequently, methods of conversion of numbers from weighted systems to...
-
Fusion-based Representation Learning Model for Multimode User-generated Social Network Content
PublicationAs mobile networks and APPs are developed, user-generated content (UGC), which includes multi-source heterogeneous data like user reviews, tags, scores, images, and videos, has become an essential basis for improving the quality of personalized services. Due to the multi-source heterogeneous nature of the data, big data fusion offers both promise and drawbacks. With the rise of mobile networks and applications, UGC, which includes...
-
FPGA implementation of the multiplication operation in multiple-precision arithmetic
PublicationAlthough standard 32/64-bit arithmetic is sufficient to solve most of the scientific-computing problems, there are still problems that require higher numerical precision. Multiple-precision arithmetic (MPA) libraries are software tools for emulation of computations in a user-defined precision. However, availability of a reconfigurable cards based on field-programmable gate arrays (FPGAs) in computing systems allows one to implement...
-
Cascade Object Detection and Remote Sensing Object Detection Method Based on Trainable Activation Function
PublicationObject detection is an important process in surveillance system to locate objects and it is considered as major application in computer vision. The Convolution Neural Network (CNN) based models have been developed by many researchers for object detection to achieve higher performance. However, existing models have some limitations such as overfitting problem and lower efficiency in small object detection. Object detection in remote...