computation speed
Recently Published Documents


TOTAL DOCUMENTS

115
(FIVE YEARS 34)

H-INDEX

9
(FIVE YEARS 1)

2022 ◽  
Vol 12 (1) ◽  
pp. 468
Author(s):  
Yeonghyeon Gu ◽  
Zhegao Piao ◽  
Seong Joon Yoo

In magnetic resonance imaging (MRI) segmentation, conventional approaches utilize U-Net models with encoder–decoder structures, segmentation models using vision transformers, or models that combine a vision transformer with an encoder–decoder model structure. However, conventional models have large sizes and slow computation speed and, in vision transformer models, the computation amount sharply increases with the image size. To overcome these problems, this paper proposes a model that combines Swin transformer blocks and a lightweight U-Net type model that has an HarDNet blocks-based encoder–decoder structure. To maintain the features of the hierarchical transformer and shifted-windows approach of the Swin transformer model, the Swin transformer is used in the first skip connection layer of the encoder instead of in the encoder–decoder bottleneck. The proposed model, called STHarDNet, was evaluated by separating the anatomical tracings of lesions after stroke (ATLAS) dataset, which comprises 229 T1-weighted MRI images, into training and validation datasets. It achieved Dice, IoU, precision, and recall values of 0.5547, 0.4185, 0.6764, and 0.5286, respectively, which are better than those of the state-of-the-art models U-Net, SegNet, PSPNet, FCHarDNet, TransHarDNet, Swin Transformer, Swin UNet, X-Net, and D-UNet. Thus, STHarDNet improves the accuracy and speed of MRI image-based stroke diagnosis.


2021 ◽  
Author(s):  
Tan Yongliang ◽  
He Lesheng ◽  
Jin Haonan ◽  
Kong Qingyang

As quantum computing and the theory of bilinear pairings continue being studied in depth, elliptic curves on GF(3m ) are becoming of an increasing interest because they provide a higher security. What’s more, because hardware encryption is more efficient and secure than software encryption in today's IoT security environment, this article implements a scalar multiplication algorithm for the elliptic curve on GF(3m ) on the FPGA device platform. The arithmetic in finite fields is quickly implemented by bit-oriented operations, and then the computation speed of point doubling and point addition is improved by a modified Jacobia projection coordinate system. The final experimental results demonstrate that the structure consumes a total of 7518 slices, which is capable of computing approximately 3000 scalar multiplications per second at 124 Mhz. It has relative advantages in terms of performance and resource consumption, which can be applied to specific confidential communication scenarios as an IP core.


Sensors ◽  
2021 ◽  
Vol 21 (20) ◽  
pp. 6927
Author(s):  
Xiaojuan Wang ◽  
Xinlei Wang ◽  
Tianqi Lv ◽  
Lei Jin ◽  
Mingshu He

Human activity recognition (HAR) based on wearable sensors is a promising research direction. The resources of handheld terminals and wearable devices limit the performance of recognition and require lightweight architectures. With the development of deep learning, the neural architecture search (NAS) has emerged in an attempt to minimize human intervention. We propose an approach for using NAS to search for models suitable for HAR tasks, namely, HARNAS. The multi-objective search algorithm NSGA-II is used as the search strategy of HARNAS. To make a trade-off between the performance and computation speed of a model, the F1 score and the number of floating-point operations (FLOPs) are selected, resulting in a bi-objective problem. However, the computation speed of a model not only depends on the complexity, but is also related to the memory access cost (MAC). Therefore, we expand the bi-objective search to a tri-objective strategy. We use the Opportunity dataset as the basis for most experiments and also evaluate the portability of the model on the UniMiB-SHAR dataset. The experimental results show that HARNAS designed without manual adjustments can achieve better performance than the best model tweaked by humans. HARNAS obtained an F1 score of 92.16% and parameters of 0.32 MB on the Opportunity dataset.


Energies ◽  
2021 ◽  
Vol 14 (20) ◽  
pp. 6651
Author(s):  
Remigiusz Iwańkowicz

This paper addresses the problem of route planning for a fleet of electric vehicles departing from a depot and supplying customers with certain goods. This paper aims to present a permutation-based method of vehicle route coding adapted to the specificity of electric drive. The developed method integrated with an evolutionary algorithm allows for rapid generation of routes for multiple vehicles taking into account the necessity of supplying energy in available charging stations. The minimization of the route distance travelled by all vehicles was taken as a criterion. The performed testing indicated satisfactory computation speed. A real region with four charging stations and 33 customers was analysed. Different scenarios of demand were analysed, and factors affecting the results of the proposed calculation method were indicated. The limitations of the method were pointed out, mainly caused by assumptions that simplify the problem. In the future, it is planned for research and method development to include the lapse of time and for the set of factors influencing energy consumption by a moving vehicle to be extended.


Over time, an exorbitant data quantity is generating which indeed requires a shrewd technique for handling such a big database to smoothen the data storage and disseminating process. Storing and exploiting such big data quantities require enough capable systems with a proactive mechanism to meet the technological challenges too. The available traditional Distributed File System (DFS) becomes inevitable while handling the dynamic variations and requires undefined settling time. Therefore, to address such huge data handling challenges, a proactive grid base data management approach is proposed which arranges the huge data into various tiny chunks called grids and makes the placement according to the currently available slots. The data durability and computation speed have been aligned by designing data disseminating and data eligibility replacement algorithms. This approach scrumptiously enhances the durability of data accessing and writing speed. The performance has been tested through numerous grid datasets and therefore, chunks have been analysed through various iterations by fixing the initial chunks statistics, then making a predefined chunk suggestion and then relocating the chunks after the substantial iterations and found that chunks are in an optimal node from the first iteration of replacement which is more than 21% of working clusters as compared to the traditional approach.


2021 ◽  
Vol 11 (16) ◽  
pp. 7741
Author(s):  
Wooryong Park ◽  
Donghee Lee ◽  
Junhak Yi ◽  
Woochul Nam

Tracking a micro aerial vehicle (MAV) is challenging because of its small size and swift motion. A new model was developed by combining compact and adaptive search region (SR). The model can accurately and robustly track MAVs with a fast computation speed. A compact SR, which is slightly larger than a target MAV, is less likely to include a distracting background than a large SR; thus, it can accurately track the MAV. Moreover, the compact SR reduces the computation time because tracking can be conducted with a relatively shallow network. An optimal SR to MAV size ratio was obtained in this study. However, this optimal compact SR causes frequent tracking failures in the presence of the dynamic MAV motion. An adaptive SR is proposed to address this problem; it adaptively changes the location and size of the SR based on the size, location, and velocity of the MAV in the SR. The compact SR without adaptive strategy tracks the MAV with an accuracy of 0.613 and a robustness of 0.086, whereas the compact and adaptive SR has an accuracy of 0.811 and a robustness of 1.0. Moreover, online tracking is accomplished within approximately 400 frames per second, which is significantly faster than the real-time speed.


2021 ◽  
Vol 15 ◽  
Author(s):  
Wooseok Choi ◽  
Myonghoon Kwak ◽  
Seyoung Kim ◽  
Hyunsang Hwang

Hardware neural network (HNN) based on analog synapse array excels in accelerating parallel computations. To implement an energy-efficient HNN with high accuracy, high-precision synaptic devices and fully-parallel array operations are essential. However, existing resistive memory (RRAM) devices can represent only a finite number of conductance states. Recently, there have been attempts to compensate device nonidealities using multiple devices per weight. While there is a benefit, it is difficult to apply the existing parallel updating scheme to the synaptic units, which significantly increases updating process’s cost in terms of computation speed, energy, and complexity. Here, we propose an RRAM-based hybrid synaptic unit consisting of a “big” synapse and a “small” synapse, and a related training method. Unlike previous attempts, array-wise fully-parallel learning is possible with our proposed architecture with a simple array selection logic. To experimentally verify the hybrid synapse, we exploit Mo/TiOx RRAM, which shows promising synaptic properties and areal dependency of conductance precision. By realizing the intrinsic gain via proportionally scaled device area, we show that the big and small synapse can be implemented at the device-level without modifications to the operational scheme. Through neural network simulations, we confirm that RRAM-based hybrid synapse with the proposed learning method achieves maximum accuracy of 97 %, comparable to floating-point implementation (97.92%) of the software even with only 50 conductance states in each device. Our results promise training efficiency and inference accuracy by using existing RRAM devices.


Author(s):  
Muhammad Abdul Haq ◽  
Iwan Kurnianto Wibowo ◽  
Bima Sena Bayu Dewantara

This paper presents a novel approach for improving the computation speed of the ball detection and obstacle detection processes in our robot. The conditions of obstacle detection and ball detection in our robot still have a slow processing speed, this condition makes the robot not real-time and the robot's movement is hampered. To build a good world model, things to note are obstacle information and real-time ball detection. The focus of this research is to increase the speed of the process of the ball and obstacle detection around the robot. To increase the speed of the process, it is necessary to optimize the use of the OpenCV library on the robot operating system (ROS) system to divide the process into several small processes so that the work can be optimally divided into threads that have been created. Then, minimize the use of too many frames. This information will be sent to the base station and also for obstacle avoidance purposes.


Information ◽  
2021 ◽  
Vol 12 (6) ◽  
pp. 228
Author(s):  
Hongbin Wang ◽  
Pengming Wang ◽  
Shengchun Deng ◽  
Haoran Li

As the classic feature selection algorithm, the Relief algorithm has the advantages of simple computation and high efficiency, but the algorithm itself is limited to only dealing with binary classification problems, and the comprehensive distinguishing ability of the feature subsets composed of the former K features selected by the Relief algorithm is often redundant, as the algorithm cannot select the ideal feature subset. When calculating the correlation and redundancy between characteristics by mutual information, the computation speed is slow because of the high computational complexity and the method’s need to calculate the probability density function of the corresponding features. Aiming to solve the above problems, we first improve the weight of the Relief algorithm, so that it can be used to evaluate a set of candidate feature sets. Then we use the improved joint mutual information evaluation function to replace the basic mutual information computation and solve the problem of computation speed and correlation, and redundancy between features. Finally, a compound correlation feature selection algorithm based on Relief and joint mutual information is proposed using the evaluation function and the heuristic sequential forward search strategy. This algorithm can effectively select feature subsets with small redundancy and strong classification characteristics, and has the excellent characteristics of faster calculation speed.


2021 ◽  
Author(s):  
Rofeide Jabbari

In this thesis we study and analyze the pricing of barrier and barrier crack options under a Time-Changed Levy process. Oil and gasoline in Canada are our underlying commodities of interest in this study. To characterize the dynamics of oil and gasoline prices, Black-Scholes and Time-Changed models based on Levy process are proposed. To verify the model, real data of the Canada oil and gas market is used. While the pricing methods based on Monte Carlo are the well-known and dominant for price calculation, we propose a Fourier Transform (FT) for the pricing, which provide some important advantages to the Monte Carlo method such as computation speed without compromising any accuracy. The method is also applied to Crack spread contracts to reduce the risk.


Sign in / Sign up

Export Citation Format

Share Document