检索结果-内蒙古大学图书馆

61st ACM/IEEE Design Automation Conference, DAC 2024

作者： Shin, Dongjin Choi, Insu Yang, Joon-Sung Department of Electrical and Electronic Engineering Yonsei University Seoul Korea Republic of Department of Systems Semiconductor Engineering Yonsei University Seoul Korea Republic of BK21 Graduate Program in Intelligent Semiconductor Technology Yonsei University Seoul Korea Republic of

ISBN: (纸本)9798400706011

Vision Transformers have demonstrated remarkable performance in various vision tasks. However, general-purpose processors, such as CPUs and GPUs, face challenges in efficiently handling the inference of Vision Transformers. To address the issue, prior works have focused on accelerating only attention due to its high computational cost in NLP Transformers. In contrast, Vision Transformers demonstrate a higher computational cost due to linear modules such as linear transformation, linear projection and Feed-Forward Network (FFN), compared to attention. In this paper, we present ViT-slice, an algorithm-architecture co-design that enhances end-to-end performance and energy efficiency by optimizing not only attention but also linear modules. At the algorithm level, we propose bit-slice compression that avoids storing the redundant most significant bits (MSBs). Additionally, we present bit-slice dot product with early skip to efficiently compute the dot product using bit-sliced data. To enable early skip during the dot product computation, we leverage a trainable threshold. On the hardware level, we introduce a specialized bit-slice dot product unit (BSDPU) to efficiently process the bit-slice dot product with early skip algorithm. Additionally, we present a bit-slice encoder and decoder for on-chip bit-slice compression. ViT-slice achieves 244×, 35.3×, 16.8×, 10.4×, 5.0× end-to-end speedup over Xeon CPU, EdgeGPU, TITAN Xp GPU, Sanger accelerator and ViTCoD accelerator, respectively. © 2024 Copyright held by the owner/author(s).

关键词： Linear transformations

来源：评论

学校读者我要写书评

暂无评论

Morphology and Alignment Transition of Hexabenzocoronene (HBC) Mesogen Films by Bar Coating: Effect of Coating Speed

引用

Langmuir 2024年第32期40卷 16846-16854页

作者： Yang, Hao-Chun Du, You-Sheng Lee, Jey-Jau Yeh, Chun-Hong Tseng, Mei-Chun Ho, Yi-Chi Kuo, Han-Wen Yoshida, Hiroyuki Fujii, Akihiko Ozaki, Masanori Tao, Yu-Tai Akutagawg, Tomoyuki Chen, Hsiu-Hui Department of Molecular Science and Engineering National Taipei University of Technology Taipei106 Taiwan National Synchrotron Radiation Research Center Hsinchu300 Taiwan Institute of Chemistry Academia Sinica Taipei115 Taiwan School of Engineering Building VII Kwansei Gakuin University Sanda662-8501 Japan Department of Electrical and Electronic Systems Engineering Osaka Institute of Technology Omiya Osaka535-8585 Japan Division of Electrical Electronic and Infocommunications Engineering Graduate School of Engineering Osaka University Osaka565-0871 Japan Graduate School of Engineering Tohoku University Sendai980-8577 Japan Tohoku University 2-1-1 Katahira Sendai980-8577 Japan

Films of the discotic liquid crystalline hexabenzocoronene (HBC) derivative, HBC-1,3,5-Ph-C12, were prepared on the quartz substrate by the bar-coating method. Depending on the coating speed, regularly spaced stripes or continuous films were observed. In the former case, columns of the HBC derivatives align more along the stripes, which are perpendicular to the coating direction, whereas in the latter case, columns of the HBC derivatives in the film align more along the coating direction. These distinctive structures are confirmed via polarized optical microscopy (POM), polarized UV-vis spectroscopy, and grazing incidence small-angle X-ray scattering measurements. © 2024 The Authors. Published by American Chemical Society

关键词： Coatings

来源：评论

学校读者我要写书评

暂无评论

SFA-Net: A Selective Features Absorption Network for Object Detection in Rainy Weather Conditions

引用

IEEE Transactions on Neural Networks and Learning Systems 2023年第8期34卷 5122-5132页

作者： Huang, Shih-Chia Hoang, Quoc-Viet Le, Trung-Hieu National Taipei University of Technology Department of Electronic Engineering Taipei10608 Taiwan National Taipei University of Technology International Graduate Program in Electrical Engineering and Computer Science Taipei10608 Taiwan Hung Yen University of Technology and Education Faculty of Information Technology Hungyen 17000 Viet Nam

In recent years, object detection approaches using deep convolutional neural networks (CNNs) have derived major advances in normal images. However, such success is hardly achieved with rainy images due to lack of visibility. Aiming to bridge this gap, in this article, we present a novel selective features absorption network (SFA-Net) to improve the performance of object detection not only in rainy weather conditions but also in favorable weather conditions. SFA-Net accomplishes this objective by utilizing three subnetworks, where the feature selection subnetwork is concatenated with the object detection subnetwork through the feature absorption subnetwork to form a unified model. To promote further advancement in object detection impaired by rain, we propose a large-scale rainy image dataset, named srRain, which contains both synthetic rainy images and real-world rainy images for training and testing purposes. srRain is comprised of 25 900 rainy images depicting diverse driving scenarios in the presence of rain with a total of 181 164 instances interpreting five common object categories. Experimental results display that our SFA-Net reaches the highest mean average precision (mAP) of 77.53% on a normal image set, 62.52% on a synthetic rainy image set, 37.34% on a collected natural rainy image set, and 32.86% on a published real rainy image set, surpassing current state-of-the-art object detectors and the combination of image deraining and object detection models while retaining a high speed. © 2012 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

PROPOSAL ON DISTRIBUTED MQTT BROKERS ARCHITECTURE 22

PROPOSAL ON DISTRIBUTED MQTT BROKERS ARCHITECTURE

引用

22nd International Conference on e-Society 2024, ES 2024 and 20th International Conference on Mobile Learning 2024, ML 2024

作者： Kosaka, Kazuhiro Mitsuuchi, Masaki Yokotani, Tetsuya Ishibashi, Koichi The Graduate School of Electrical Engineering and Electronics Kanazawa Institute of Technology Ishikawa Nonoichi Japan Department of Electrical and Electronic Engineering College of Engineering Kanazawa Institute of Technology Ishikawa Nonoichi Japan Department of Information and Computer Science College of Engineering Kanazawa Institute of Technology Ishikawa Hakusan Japan

Message Queue Telemetry Transport (MQTT) is a popular communication protocol for IoT, in which Publishers and Subscribers communicate through a single broker. Therefore, a distributed broker is necessary to accommodate multiple Publishers and Subscribers at multiple locations. Previous studies have used a method in which multiple brokers are connected in a ring to share information. However, in this method, the processing delay time increases as the number of brokers and Hops increases. In this study, we propose a direct connection method between distributed brokers that reduces the processing delay time. We also use an Analytic Hierarchy Process (AHP) to determine the communication protocol used to connect brokers. Next, partial prototyping was performed to estimate the latency. © 2024 International Conferences e-Society 2024 and Mobile Learning 2024. All rights reserved.

关键词： DDS Distributed Brokers IoT MQTT

来源：评论

学校读者我要写书评

暂无评论

Efficient Remote Identification for Drone Swarms

Efficient Remote Identification for Drone Swarms

引用

作者： Seo, Kang-Moon Kim, Jane Lee, Soojin Kwon, Jun-Woo Seo, Seung-Hyun The Department of Electronic & Electrical Engineering Graduate School Hanyang University Ansan Korea Republic of School of Electrical Engineering Hanyang University ERICA Ansan Korea Republic of

With the advancement of unmanned aerial vehicle (UAV) technology, the market for drones and the cooperation of many drones are expanding. Drone swarms move together in multiple regions to perform their tasks. A Ground Control Server (GCS) located in each region identifies drone swarm members to prevent unauthorized drones from trespassing. Studies on drone identification have been actively conducted, but existing studies did not consider multiple drone identification environments. Thus, developing a secure and effective identification mechanism for drone swarms is necessary. We suggested a novel approach for the remote identification of drone swarms. For an efficient identification process between the drone swarm and the GCS, each Reader drone in the region collects the identification information of the drone swarm and submits it to the GCS for verification. The proposed identification protocol reduces the verification time for a drone swarm by utilizing batch verification to verify numerous drones in a drone swarm simultaneously. To prove the security and correctness of the proposed protocol, we conducted a formal security verification using ProVerif, an automatic cryptographic protocol verifier. We also implemented a non-flying drone swarm prototype using multiple Raspberry Pis to evaluate the proposed protocol’s computational overhead and effectiveness. We showed simulation results regarding various drone simulation scenarios. © 2023 Tech Science Press. All rights reserved.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

Magnetic Absolute Encoder with Magnet of Different Magnetic Flux Densities and AMR Sensors 49

Magnetic Absolute Encoder with Magnet of Different Magnetic ...

引用

49th Annual Conference of the IEEE Industrial electronics Society, IECON 2023

作者： Sakuma, Yusuke Takashima, Kohei Ohhira, Takashi Hashimoto, Hideki Graduate School of Science and Engineering Chuo University Tokyo Japan Chuo University Department of Electrical Electronic and Communication Engineering Tokyo Japan

ISBN: (纸本)9798350331820

In this study, we developed a novel magnetic absolute encoder with different magnetic flux densities. Magnetic encoders can be created with a simple structure consisting of a magnet magnetized in a sinusoidal shape and a sensor that detects the magnetic field. They are known for their compact design and excellent environmental resistance. A method of obtaining even higher resolution signals while maintaining absolute accuracy is to use a magnet with different magnetic flux densities and Hall sensors. However, there are physical limitations in improving the accuracy of a multi-polarizing magnet with different magnetic flux densities. In this study, we developed a magnetic encoder with a new structure using a magnet with different magnetic flux densities and anisotropic magnetoresistive(AMR) sensors. The developed system can calculate the absolute angles and output angular frequency signals twice as high as those of AMR sensors, thereby improving the resolution without requiring a multipole magnet. In the past, distortion was considered to occur because AMR sensors were used outside the saturation sensitivity region to capture the amplitude characteristics of a magnet with different magnetic flux densities;however, this was solved by correction using a look-up table. Finally, an actual device was created and validated by comparing the measured values obtained in the experiment with those obtained using a Hall sensor. © 2023 IEEE.

关键词： Magnetic flux

来源：评论

学校读者我要写书评

暂无评论

Generation of Synthetic Series for Hydroelectric Plants in Brazilian Interconnected System by Metaheuristic 59

Generation of Synthetic Series for Hydroelectric Plants in B...

引用

59th International Universities Power engineering Conference, UPEC 2024

作者： Mosqueira Furtado, Renato Andrade Marques Marcato, Andre Luis Da Silva Junior, Ivo Chaves Federal University of Juiz de Fora Electrical Engineering Graduate Program Juiz de Fora Brazil Federal University of Juiz de Fora Department of Electrical Engineering Juiz de Fora Brazil

ISBN: (纸本)9798350379730

This article aims to explain the use of the intelligent technique known as Arithmetic Optimization Algorithm for generating synthetic series for hydroelectric plants which belong to the Brazilian interconnected system. Thus, the metaheuristic is responsible for determining the best coefficients as well as the optimal order for each month which may represent historical data based on an optimization problem whose objective function is to minimize the standard deviation of errors between the real and estimated water inflow values. Finally, with the optimal coefficients and order, the next step is to generate synthetic flow series considering a period of five years ahead, guaranteeing forecasts of inflow energy which are coherent with the historical data. The simulations were carried out considering Itaipu, Belo Monte, Tucuruí and Comp PAF-MOX hydroelectric plants whose results show the applicability of the Arithmetic Optimization Algorithm for generating flow scenarios allowing a convincing energy planning. © 2024 IEEE.

关键词： Hydroelectric power plants

来源：评论

学校读者我要写书评

暂无评论

Maneuverability Improvement by Feedback Compensation of a Physically Human-Interacting Robot with Variable Damping Control

Maneuverability Improvement by Feedback Compensation of a Ph...

引用

18th IEEE International Conference on Advanced Motion Control, AMC 2024

作者： Songthumjitti, Narawich Inara, Takpeshi Graduate School of Tokai University Course of Electrical and Electronic Engineering Kanagawa Japan Tokai University Department of Applied Computer Engineering Kanagawa Japan

ISBN: (纸本)9798350342819

Manufacturing workflows can be enhanced using the implementation of robots;however, it is limited to repetitive tasks, and a human operator is still needed to gain flexibility in the robot's movement. However, human-robot interaction can cause instability due to the robot's characteristics. The typical solution is to decrease the admittance parameter, but it also reduces the assisting effectiveness. Variable damping is another way to counter this problem, but if the instability problem isn't solved yet, the effect can be difficult to perceive. Our study aims to implement feedback compensation in the robot system to reduce the effect of actuator dead time and insufficient structural stiffness. The results prove that the implementation of feedback compensation can greatly improve the maneuverability of the robot by decreasing the force required during the traveling phase and increasing accuracy in the positioning phase, resulting in a task that can be completed with less effort and a shorter operation time. © 2024 IEEE.

关键词： Human robot interaction

来源：评论

学校读者我要写书评

暂无评论

Adaptive Fault Detection Based on Neural Networks and Multiple Sampling Points for Distribution Networks and Microgrids

引用

Journal of Modern Power Systems and Clean Energy 2022年第6期10卷 1648-1657页

作者： J.Marín-Quintero C.Orozco-Henao A.S.Bretas J.C.Velez A.Herrada A.Barranco-Carlos W.S.Percybrooks Electric and Electronic Engineering Program Universidad Tecnológica de BolívarCartagenaColombia Electrical and Electronic Engineering Department Universidad del NorteBarranquillaColombia Distributed Systems Group Pacific Northwest National LaboratoryGainesvilleUSA

Smart networks such as active distribution network(ADN)and microgrid(MG)play an important role in power system *** design and implementation of appropriate protection systems for MG and ADN must be addressed,which imposes new technical *** paper presents the implementation and validation aspects of an adaptive fault detection strategy based on neural networks(NNs)and multiple sampling points for ADN and *** solution is implemented on an edge *** are used to derive a data-driven model that uses only local measurements to detect fault states of the network without the need for communication *** sampling points are used to derive a data-driven model,which allows the generalization considering the implementation in physical *** adaptive fault detector model is implemented on a Jetson Nano system,which is a single-board computer(SBC)with a small graphic processing unit(GPU)intended to run machine learning loads at the *** proposed method is tested in a physical,real-life,low-voltage network located at Universidad del Norte,*** testing network is based on the IEEE 13-node test feeder scaled down to 220 *** validation in a simulation environment shows the accuracy and dependability above 99.6%,while the real-time tests show the accuracy and dependability of 95.5%and 100%,*** hard-to-derive parameters,the easy-to-implement embedded model highlights the potential for real-life applications.

关键词： Distribution network microgrid adaptive protection intelligent electronic device machine learning

来源：评论

学校读者我要写书评

暂无评论

Degradation of Organic LEDs Characterized with Algebraic Decay Relaxation Function 15

Degradation of Organic LEDs Characterized with Algebraic Dec...

引用

Advances in Display Technologies XV 2025

作者： Aygen, Can Richards, Michael Yazdaanpanah, Ramin Kulkarni, Sachin Tanriover, Ibrahim Onsager, Claire Lin, BoYen Lee, Jiun-Haw Aydin, Koray Grayson, Matthew A. Department of Electrical & Computer Engineering Northwestern University EvanstonIL United States Department of Opto-Electronic Engineering National Dong Hwa University Hualien Taiwan Graduate Institute of Photonics and Optoelectronics Department of Electrical Engineering National Taiwan University Taipei Taiwan Program in Applied Physics Northwestern University EvanstonIL United States

ISBN: (纸本)9781510685246

Lifetime studies of organic light emitting diodes are necessary to both understand and mitigate causes of degradation. Here, electroluminescence (EL) studies of green organic light-emitting diode (OLED) pixels are measured continuously using three different dc currents over a period of one week, and their slow EL degradation is accurately fit to an analytical expression with only two fit parameters. The slower-than-exponential EL decay is fit to an empirical unified heavy-tail function that can interpolate between power-law algebraic decays (AD) and stretched-exponential decays (SE). Whereas SE and AD are largely indistinguishable down to 50% of the initial luminance, here we report degradation down to 10% of the initial luminance which most closely fits an AD function with remarkable precision over the entire duration of the experiment. All relaxations share the same characteristic exponent β = 0.53 in the power law decay, whereas the characteristic time-scale τ for degradation decreases with a power-law exponent B = 1.9 as the current increases. The high degree of accuracy in the fit allows asymptotic predictions at time scales that are orders of magnitude longer than the duration of the data collection. The voltage across the OLED is simultaneously monitored under constant-current degradation and reveals an early-onset time scale where the voltage reaches a minimum before degradation runoff. © 2025 SPIE.

关键词： Organic light emitting diodes (OLED)

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：