Heterogeneous computing systems require an efficient way of distributing tasks across processing nodes. The tasks have to be mapped to the processors which execute them in the shortest time possible, while keeping the...
详细信息
Heterogeneous computing systems require an efficient way of distributing tasks across processing nodes. The tasks have to be mapped to the processors which execute them in the shortest time possible, while keeping the processors at a similar load. Tests have shown that, in most cases, the genetic algorithm produces the best solution among all the mapping heuristics. This paper presents a Genetic Algorithm with a 3-Step Mutation which significantly increases the solution's convergence rate by using a combination of methods to mutate a chromosome. Beside the standard random approach, we implemented a targeted mutation operator which lightens the load of the most occupied processors. We also focused on different fitness functions in order to improve both the makespan and the load balance. The mutation combinations and the fitness functions are then tested to see which ones perform better and in what cases.
Evolutionary optimization plays an important role in the representation of information from data sets originating from various technical fields and natural sciences, as it helps to explore parameter spaces for meaning...
详细信息
ISBN:
(数字)9798350368130
ISBN:
(纸本)9798350368147
Evolutionary optimization plays an important role in the representation of information from data sets originating from various technical fields and natural sciences, as it helps to explore parameter spaces for meaningful representations. Quality-Diversity (QD) methods, notably MAP-Elites variants, prove effective in diverse fields, emphasizing their ability to provide sets of high-performing solutions. This review discusses challenges in single- and multi-objective optimization, with applications in multiple directions such as image processing, visualization, and medical imaging. It also reviews QD algorithms and highlights advancements in algorithmic adaptations, user-driven optimization and the potential to explore complex feature spaces. The presented works contribute to understanding and applying evolutionary optimization for solving visualization and feature space exploration problems in various domains.
Representing relevant information from volume data sets is a problem often faced in visualization. Generating meaningful images from highly-complex volume data sets is a challenging, tedious task requiring specialized...
Representing relevant information from volume data sets is a problem often faced in visualization. Generating meaningful images from highly-complex volume data sets is a challenging, tedious task requiring specialized knowledge of the distribution and properties of the data. Traditionally, this task has been carried out manually via specialized user interfaces. We propose a volume visualization pipeline which facilitates the automatic generation of high-quality images from volume data sets. Our method involves a direct volume renderer which generates images from volume data based on visual mappings provided by a transfer function. Central to our approach is a quality-focused descriptor which exploits the properties of the distribution of gradient orientations of an alpha-bounded surface within the volume. This feature is useful for determining transfer functions that result in the rendering of corresponding images depicting various details from the volume. We show that by using this feature as an optimization objective, the generation of high quality images can be automated. Using simple genetic algorithms, we can automatically generate sets of images illustrating coherent, easily-distinguishable and high-quality surfaces of relevant structures from volume data.
The paper presents a new numerical integration routine BGKODE_DSP. This routine is based on BGKODE routine, which forms a practical application of the Kroghpsilas algorithm. BGKODE_DSP routine realizes the predictor-c...
详细信息
The paper presents a new numerical integration routine BGKODE_DSP. This routine is based on BGKODE routine, which forms a practical application of the Kroghpsilas algorithm. BGKODE_DSP routine realizes the predictor-corrector method and has been simplified to ensure its applicability in real time systems. The results of the comparison between the selected common numerical integration methods and the new routine is presented. The scope of this research also involves the range of accuracy of the solution resulting from the shorter computation time. Experiments have been carried out on experimental drive system with induction motor with the use of DTC method.
Depth information from a captured scene can provide a more complete description of objects which can be exploited for recognition purposes. In this paper we propose a new approach to static hand gesture recognition th...
详细信息
Depth information from a captured scene can provide a more complete description of objects which can be exploited for recognition purposes. In this paper we propose a new approach to static hand gesture recognition that explores only depth information acquired with a Microsoft Kinect camera. Firstly, the regions of interest are extracted from depth data and a 3D point cloud is created taking into consideration the capture device settings. Then, for each 3D point in the simplified point cloud the local spin image descriptor is computed. The performance of hand pose detection greatly depends on the 3D shape descriptor used. KPCA is used for dimensionality reduction and for noise elimination from obtained spin image histograms. The most relevant features in terms of principal components are served as input for a SVM classifier. Experimental results show a high level of accuracy in recognizing static hand gestures using spin images in 3D point clouds.
A novel protocol under a priority time constraint for real-time communication systems is proposed. This protocol can maintain two priority classes. The characteristics of the proposed protocol and its features are pre...
详细信息
A novel protocol under a priority time constraint for real-time communication systems is proposed. This protocol can maintain two priority classes. The characteristics of the proposed protocol and its features are presented, and a simulator is built to examine its performance. The performance of the protocol is compared with that of the virtual-time carrier sense multiple-access (VT-CSMA) protocol. The results show that, although the proposed protocol has less throughput than the VT-CSMA protocol, its behavior is better than that of VT-CSMA with respect to high-priority packets. This behavior is the main goal of the protocol, which reduces high-priority-packet loss.< >
The spike-timing-dependent-plasticity (STDP) is a mechanism for adjusting the efficacies of biological synapses that was observed and studied in vitro. However, the STDP effect for natural neurons in vivo is subject o...
详细信息
The spike-timing-dependent-plasticity (STDP) is a mechanism for adjusting the efficacies of biological synapses that was observed and studied in vitro. However, the STDP effect for natural neurons in vivo is subject of debate because in several experiments when the neurons are stimulated indirectly by natural paths the STDP effect was insignificant. Starting from these aspects this work studies by simulation the STDP long-term effect on synaptic plasticity in order to determine whether the long-term potentiation (LTP) and the long-term depression (LTD) could compensate each other during long-term activity of the neural network. The results show that for some synapses the weights start to oscillate in small intervals around long term stable values that are different from the limits of the weights variation interval. This behavior is caused by the fact that, indeed, the effects of LTP and LTD compensate each other at certain weight values when the same pattern of the input stimuli is presented repeatedly to the network input. The LTP and LTD compensation that determines long term weights stability to other values than the weight variation limits could improve the sensitivity of the learning process in the biological networks because no neuron specific limitation is introduced.
Wireless sensors networks is an active research topic. The sensor nodes (i.e. motes) are the main building blocks of these networks. There is a permanent concern for building more and more efficient motes in order to ...
详细信息
Wireless sensors networks is an active research topic. The sensor nodes (i.e. motes) are the main building blocks of these networks. There is a permanent concern for building more and more efficient motes in order to satisfy the demanding specifications of a sensor network. This paper introduces a new mote device: aceMOTE which is based on a 32 bit microcontroller. The hardware structure of the mote is presented and a comparison is made with similar popular devices. An experimental analysis is made regarding the energy consumption and the results prove that a 32 bit microcontroller is a viable choice for usage in mote design compared with the more frequently used 8 bit and 16 bit microcontrollers.
Summary form only given. A new scheme for transmitting sensitive data is proposed, the proposed scheme depends on partitioning the output of a block encryption module using the Chinese Remainder Theorem among a set of...
详细信息
Summary form only given. A new scheme for transmitting sensitive data is proposed, the proposed scheme depends on partitioning the output of a block encryption module using the Chinese Remainder Theorem among a set of channels. The purpose of using the Chinese Remainder Theorem is to hide the cipher text in order to increase the difficulty of attacking the cipher. The proposed scheme assumes the existence of a set of transmission channels between the sender and the receiver parties from which S channels are chosen using some selection criteria. The message is divided into as set of blocks of a size that is a multiple of the size of the underlying block cipher block size. The resulting N bit encrypted block is used as an N-bits integer. This integer remainders with respect to S selected relatively prime numbers are sent over the S selected channels. On the receiver side the inverse Chinese remainder is applied to get the encrypted blocks followed by a decryption operation to construct the original message. The theory, implementation and the security of this scheme are described in the paper.
This work presents an image vectorization application aimed to exploit features found in most of today's modern architectures. Design decisions supported by performance results are presented along with use cases f...
详细信息
ISBN:
(纸本)9781479914920
This work presents an image vectorization application aimed to exploit features found in most of today's modern architectures. Design decisions supported by performance results are presented along with use cases for the proposed solution. We will present the algorithm, the target architectures and threading models, as well as a number of optimization techniques. Finally results of both low power SoC and high performance multicore processors are discussed.
暂无评论