Computers & Electrical Engineering最新文献_第6页

A visual cortex-inspired edge neuromorphic hardware architecture with on-chip multi-layer STDP learning 受视觉皮层启发的边缘神经形态硬件架构，具有片上多层 STDP 学习功能

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-29 DOI: 10.1016/j.compeleceng.2024.109806

Junxian He , Min Tian , Ying Jiang , Haibing Wang , Tengxiao Wang , Xichuan Zhou , Liyuan Liu , Nanjian Wu , Ying Wang , Cong Shi

The era of artificial intelligence of things (AIoT) brings huge challenges on edge visual processing systems under strict processing latency, cost and energy budgets. The emergence of computationally efficient biological spiking neural networks (SNN) and event-driven neuromorphic architecture in recent years have fostered a computing paradigm shift to address these challenges. In this paper, we propose a neuromorphic processor architecture for a multi-layer convolutional SNN (codenamed HMAX SNN model) inspired by human visual cortex hierarchy. The main contributions of this work include: 1) It proposes a fully event-driven, modular, configurable and scalable neuromorphic architecture allowing for flexible tradeoffs among implementation cost, processing speed and visual recognition accuracy with multi-layer convolutional SNNs. 2) It proposes a run-time reconfigurable learning engine enabling fast on-chip unsupervised spike-timing dependent plasticity (STDP) learning for the feature-extraction convolutional layers and also supervised STDP learning for the feature-classification FC layer, in a time-multiplexing way. These techniques greatly improve on-chip learning accuracies beyond 97 % on the Modified National Institute of Standards and Technology database (MNIST) images for the first time among existing edge neuromorphic systems, at reasonable computational and memory costs. Our hardware processor architecture was prototyped on a low-cost Zedboard Zynq-7020 Field-Programmable Gate Array (FPGA) device, and validated on the MNIST, Fashion-MNIST, Olivetti Research Laboratory (ORL) human faces and ETH-80 image datasets. The experimental results demonstrate that the proposed neuromorphic architecture can achieve comparably high on-chip learning accuracy, high inference throughput and high energy efficiency using relatively fewer hardware resource consumptions. We anticipate that the HMAX SNN processor can potentially enhance deployments of edge neuromorphic processors in more practical edge applications.

人工智能物联网（AIoT）时代在严格的处理延迟、成本和能耗预算下给边缘视觉处理系统带来了巨大挑战。近年来，计算高效的生物尖峰神经网络（SNN）和事件驱动神经形态架构的出现，促进了计算模式的转变，以应对这些挑战。在本文中，我们受人类视觉皮层层次结构的启发，为多层卷积 SNN（代号为 HMAX SNN 模型）提出了一种神经形态处理器架构。这项工作的主要贡献包括1) 它提出了一种完全事件驱动、模块化、可配置和可扩展的神经形态架构，允许在多层卷积 SNN 的实施成本、处理速度和视觉识别准确性之间灵活权衡。2) 它提出了一种运行时可重新配置的学习引擎，能以时间多路复用的方式，对特征提取卷积层进行快速的片上无监督尖峰计时可塑性（STDP）学习，并对特征分类 FC 层进行有监督 STDP 学习。在现有的边缘神经形态系统中，这些技术以合理的计算和内存成本，首次将修改后的美国国家标准与技术研究院（National Institute of Standards and Technology）数据库（MNIST）图像的片上学习准确率大大提高到 97% 以上。我们在低成本的 Zedboard Zynq-7020 现场可编程门阵列（FPGA）设备上搭建了硬件处理器架构原型，并在 MNIST、Fashion-MNIST、Olivetti 研究实验室（ORL）人脸和 ETH-80 图像数据集上进行了验证。实验结果表明，所提出的神经形态架构能以相对较少的硬件资源消耗实现相当高的片上学习精度、高推理吞吐量和高能效。我们预计，HMAX SNN 处理器有可能在更多实际边缘应用中加强边缘神经形态处理器的部署。

{"title":"A visual cortex-inspired edge neuromorphic hardware architecture with on-chip multi-layer STDP learning","authors":"Junxian He , Min Tian , Ying Jiang , Haibing Wang , Tengxiao Wang , Xichuan Zhou , Liyuan Liu , Nanjian Wu , Ying Wang , Cong Shi","doi":"10.1016/j.compeleceng.2024.109806","DOIUrl":"10.1016/j.compeleceng.2024.109806","url":null,"abstract":"<div><div>The era of artificial intelligence of things (AIoT) brings huge challenges on edge visual processing systems under strict processing latency, cost and energy budgets. The emergence of computationally efficient biological spiking neural networks (SNN) and event-driven neuromorphic architecture in recent years have fostered a computing paradigm shift to address these challenges. In this paper, we propose a neuromorphic processor architecture for a multi-layer convolutional SNN (codenamed HMAX SNN model) inspired by human visual cortex hierarchy. The main contributions of this work include: 1) It proposes a fully event-driven, modular, configurable and scalable neuromorphic architecture allowing for flexible tradeoffs among implementation cost, processing speed and visual recognition accuracy with multi-layer convolutional SNNs. 2) It proposes a run-time reconfigurable learning engine enabling fast on-chip unsupervised spike-timing dependent plasticity (STDP) learning for the feature-extraction convolutional layers and also supervised STDP learning for the feature-classification FC layer, in a time-multiplexing way. These techniques greatly improve on-chip learning accuracies beyond 97 % on the Modified National Institute of Standards and Technology database (MNIST) images for the first time among existing edge neuromorphic systems, at reasonable computational and memory costs. Our hardware processor architecture was prototyped on a low-cost Zedboard Zynq-7020 Field-Programmable Gate Array (FPGA) device, and validated on the MNIST, Fashion-MNIST, Olivetti Research Laboratory (ORL) human faces and ETH-80 image datasets. The experimental results demonstrate that the proposed neuromorphic architecture can achieve comparably high on-chip learning accuracy, high inference throughput and high energy efficiency using relatively fewer hardware resource consumptions. We anticipate that the HMAX SNN processor can potentially enhance deployments of edge neuromorphic processors in more practical edge applications.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"120 ","pages":"Article 109806"},"PeriodicalIF":4.0,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142539641","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Applications of artificial intelligence and LiDAR in forest inventories: A Systematic Literature Review 人工智能和激光雷达在森林资源调查中的应用：系统文献综述

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-29 DOI: 10.1016/j.compeleceng.2024.109793

Welington G. Rodrigues , Gabriel S. Vieira , Christian D. Cabacinha , Renato F. Bulcão-Neto , Fabrizzio Soares

Forest inventory is a crucial tool for managing forest resources by providing quantitative and qualitative information about a particular region, much of which is collected manually in the field. Using devices such as Light Detection and Ranging (LiDAR) assists in collecting and analyzing various parameters of forest inventory. Adopting artificial intelligence (AI) techniques has sparked interest among forestry engineers seeking to work with forest LiDAR data. In this context, this study presents a Systematic Literature Review (SLR) to identify, evaluate, and interpret the results of primary studies related to the intersection between AI and Forestry Engineering. The automated search strategy retrieved 218 studies, of which 46 were selected after applying inclusion and exclusion criteria and quality assessment. After analyzing and synthesizing the data, the results showed that deep learning is becoming an increasing trend in recent research and that the direct estimation of tree diameter from aerial scans, although critical, has been minimally explored, highlighting an open field for future research.

森林资源清查是管理森林资源的重要工具，可提供特定区域的定量和定性信息，其中大部分信息都是在实地人工收集的。使用光探测和测距（LiDAR）等设备有助于收集和分析森林资源调查的各种参数。采用人工智能（AI）技术引发了林业工程师对使用森林 LiDAR 数据的兴趣。在此背景下，本研究通过系统文献综述（SLR）来识别、评估和解释与人工智能和林业工程交叉相关的主要研究成果。自动搜索策略检索到 218 项研究，在应用纳入和排除标准并进行质量评估后，从中筛选出 46 项研究。在对数据进行分析和综合后，结果表明，深度学习在近期的研究中日益成为一种趋势，而从航空扫描中直接估算树木直径虽然至关重要，但却很少得到探索，这凸显了未来研究的一个开放领域。

{"title":"Applications of artificial intelligence and LiDAR in forest inventories: A Systematic Literature Review","authors":"Welington G. Rodrigues , Gabriel S. Vieira , Christian D. Cabacinha , Renato F. Bulcão-Neto , Fabrizzio Soares","doi":"10.1016/j.compeleceng.2024.109793","DOIUrl":"10.1016/j.compeleceng.2024.109793","url":null,"abstract":"<div><div>Forest inventory is a crucial tool for managing forest resources by providing quantitative and qualitative information about a particular region, much of which is collected manually in the field. Using devices such as Light Detection and Ranging (LiDAR) assists in collecting and analyzing various parameters of forest inventory. Adopting artificial intelligence (AI) techniques has sparked interest among forestry engineers seeking to work with forest LiDAR data. In this context, this study presents a Systematic Literature Review (SLR) to identify, evaluate, and interpret the results of primary studies related to the intersection between AI and Forestry Engineering. The automated search strategy retrieved 218 studies, of which 46 were selected after applying inclusion and exclusion criteria and quality assessment. After analyzing and synthesizing the data, the results showed that deep learning is becoming an increasing trend in recent research and that the direct estimation of tree diameter from aerial scans, although critical, has been minimally explored, highlighting an open field for future research.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"120 ","pages":"Article 109793"},"PeriodicalIF":4.0,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142539748","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

MA-SPRNet: A multiple attention mechanisms-based network for self-piercing riveting joint defect detection MA-SPRNet：基于多重注意机制的自冲铆接缺陷检测网络

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-29 DOI: 10.1016/j.compeleceng.2024.109798

Peng Zhang , Lun Zhao , Yu Ren , Dong Wei , Sandy To , Zeshan Abbas , Md Shafiqul Islam

Efficient detection of defects in riveted joints during the self-piercing riveting (SPR) process will help improve riveting quality. Due to the complexity of SPR defects under actual working conditions, it is difficult for traditional visual technology to detect the forming quality of SPR joints effectively. To detect SPR defects and improve the efficiency of SPR joint forming quality, we proposed a defect detection model based on a multi-attention mechanism, named Multiple Attention Self-Piercing Riveting Network (MA-SPRNet), for the detection of SPR defects. Specifically, to alleviate problems such as unclear object features in complex environments, a multi-level fusion enhancement network (MFEN) is constructed. It fuses features into each level and improves the fusion effect by adding more levels of features. In addition, to alleviate the information redundancy generated during the feature fusion process, the triple attention module (TRAM) and the efficient multi-scale attention module (EMAM) were introduced to enhance the attention of the network to SPR defective. These modules are designed to refine the attention of the network, ensuring a more targeted analysis of riveting features. In addition, the Wise Intersection over Union (WIoU) loss function is introduced, aiming to guide the network to characterize features within the region of interest and to enhance the accurate positioning of riveting defects by the network. Finally, to verify the performance of the MA-SPRNet, an SPR defect dataset was constructed, and a series of experiments based on this dataset were conducted. The detection

m A P_{0.5}

of MA-SPRNet was 82.83%. The results of experiments show that MA-SPRNet effectively realizes the detection of riveted joint defects.

有效检测自冲铆接（SPR）过程中铆接接头的缺陷有助于提高铆接质量。由于实际工况下 SPR 缺陷的复杂性，传统的视觉技术难以有效检测 SPR 接头的成型质量。为了检测 SPR 缺陷，提高 SPR 接头成形质量的效率，我们提出了一种基于多注意机制的缺陷检测模型，命名为多注意自穿刺铆接网络（MA-SPRNet），用于检测 SPR 缺陷。具体来说，为缓解复杂环境中物体特征不清晰等问题，构建了多级融合增强网络（MFEN）。它将特征融合到每个层次，并通过增加更多层次的特征来提高融合效果。此外，为了减轻特征融合过程中产生的信息冗余，还引入了三重注意力模块（TRAM）和高效多尺度注意力模块（EMAM），以增强网络对 SPR 缺陷的注意力。这些模块旨在细化网络的注意力，确保对铆接特征的分析更具针对性。此外，还引入了 Wise Intersection over Union（WIoU）损失函数，旨在引导网络分析感兴趣区域内的特征，提高网络对铆接缺陷的准确定位。最后，为了验证 MA-SPRNet 的性能，构建了 SPR 缺陷数据集，并基于该数据集进行了一系列实验。MA-SPRNet 的检测 mAP0.5 为 82.83%。实验结果表明，MA-SPRNet 有效地实现了对铆接缺陷的检测。

{"title":"MA-SPRNet: A multiple attention mechanisms-based network for self-piercing riveting joint defect detection","authors":"Peng Zhang , Lun Zhao , Yu Ren , Dong Wei , Sandy To , Zeshan Abbas , Md Shafiqul Islam","doi":"10.1016/j.compeleceng.2024.109798","DOIUrl":"10.1016/j.compeleceng.2024.109798","url":null,"abstract":"<div><div>Efficient detection of defects in riveted joints during the self-piercing riveting (SPR) process will help improve riveting quality. Due to the complexity of SPR defects under actual working conditions, it is difficult for traditional visual technology to detect the forming quality of SPR joints effectively. To detect SPR defects and improve the efficiency of SPR joint forming quality, we proposed a defect detection model based on a multi-attention mechanism, named Multiple Attention Self-Piercing Riveting Network (MA-SPRNet), for the detection of SPR defects. Specifically, to alleviate problems such as unclear object features in complex environments, a multi-level fusion enhancement network (MFEN) is constructed. It fuses features into each level and improves the fusion effect by adding more levels of features. In addition, to alleviate the information redundancy generated during the feature fusion process, the triple attention module (TRAM) and the efficient multi-scale attention module (EMAM) were introduced to enhance the attention of the network to SPR defective. These modules are designed to refine the attention of the network, ensuring a more targeted analysis of riveting features. In addition, the Wise Intersection over Union (WIoU) loss function is introduced, aiming to guide the network to characterize features within the region of interest and to enhance the accurate positioning of riveting defects by the network. Finally, to verify the performance of the MA-SPRNet, an SPR defect dataset was constructed, and a series of experiments based on this dataset were conducted. The detection <span><math><mrow><mi>m</mi><mi>A</mi><msub><mrow><mi>P</mi></mrow><mrow><mn>0</mn><mo>.</mo><mn>5</mn></mrow></msub></mrow></math></span> of MA-SPRNet was 82.83%. The results of experiments show that MA-SPRNet effectively realizes the detection of riveted joint defects.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"120 ","pages":"Article 109798"},"PeriodicalIF":4.0,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142539749","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Advancing sustainable development: Introducing a novel fast charging technique for Li-ion batteries with supercapacitor integration 推动可持续发展：为集成超级电容器的锂离子电池引入新型快速充电技术

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-28 DOI: 10.1016/j.compeleceng.2024.109810

R Hema, Venkatarangan M J

Li-ion batteries play a vital role in today's world for their wide range of applications in electrical vehicles, mobile phones, electronic gadgets etc. Applying fast charging to Li-ion batteries is crucial for increasing consumer convenience, productivity, energy efficiency, and stimulating technical innovation across various industries. With the wide-spread usage of Li-ion batteries, it is critical to consider the environmental degradation caused by their decomposition. Therefore, solutions for fast charging Li-ion batteries must not only be reliable but also sustainable to meet future demands. This paper presents a novel Li-ion battery fast charging technology with an integrated supercapacitor and battery, for quick charging without shortening the battery's life cycle. The advantage of the proposed charger is that the supercapacitor assists the Li-ion battery while charging, thereby fast charging and improving the efficiency and longevity of the battery. The utilization of supercapacitor-assisted charging diminishes the burden on lithium-ion batteries, enhancing their lifespan and consequently reducing environmental harm. To validate the performance of the proposed charging technology, an experimental investigation has been carried out. The obtained results through the investigation show the increased performance of 82 % efficiency and 40 % longevity more than the conventional method. The proposed supercapacitor-based fast charging technique is not only efficient but also sustainable and reliable.

锂离子电池在电动汽车、移动电话和电子产品等领域应用广泛，在当今世界发挥着至关重要的作用。在锂离子电池中应用快速充电技术对于提高消费者的便利性、生产率和能源效率以及促进各行业的技术创新至关重要。随着锂离子电池的广泛使用，必须考虑到其分解造成的环境恶化问题。因此，为锂离子电池快速充电的解决方案不仅要可靠，还要具有可持续性，以满足未来的需求。本文介绍了一种新型锂离子电池快速充电技术，该技术集成了超级电容器和电池，可在不缩短电池生命周期的情况下实现快速充电。所提出的充电器的优势在于，超级电容器在充电时可辅助锂离子电池，从而实现快速充电并提高电池的效率和寿命。利用超级电容器辅助充电可减轻锂离子电池的负担，延长其使用寿命，从而减少对环境的危害。为了验证所提出的充电技术的性能，我们开展了一项实验研究。调查结果显示，与传统方法相比，效率提高了 82%，寿命延长了 40%。所提出的基于超级电容器的快速充电技术不仅高效，而且可持续、可靠。

{"title":"Advancing sustainable development: Introducing a novel fast charging technique for Li-ion batteries with supercapacitor integration","authors":"R Hema, Venkatarangan M J","doi":"10.1016/j.compeleceng.2024.109810","DOIUrl":"10.1016/j.compeleceng.2024.109810","url":null,"abstract":"<div><div>Li-ion batteries play a vital role in today's world for their wide range of applications in electrical vehicles, mobile phones, electronic gadgets etc. Applying fast charging to Li-ion batteries is crucial for increasing consumer convenience, productivity, energy efficiency, and stimulating technical innovation across various industries. With the wide-spread usage of Li-ion batteries, it is critical to consider the environmental degradation caused by their decomposition. Therefore, solutions for fast charging Li-ion batteries must not only be reliable but also sustainable to meet future demands. This paper presents a novel Li-ion battery fast charging technology with an integrated supercapacitor and battery, for quick charging without shortening the battery's life cycle. The advantage of the proposed charger is that the supercapacitor assists the Li-ion battery while charging, thereby fast charging and improving the efficiency and longevity of the battery. The utilization of supercapacitor-assisted charging diminishes the burden on lithium-ion batteries, enhancing their lifespan and consequently reducing environmental harm. To validate the performance of the proposed charging technology, an experimental investigation has been carried out. The obtained results through the investigation show the increased performance of 82 % efficiency and 40 % longevity more than the conventional method. The proposed supercapacitor-based fast charging technique is not only efficient but also sustainable and reliable.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"120 ","pages":"Article 109810"},"PeriodicalIF":4.0,"publicationDate":"2024-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142526694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An elevator door anomaly detection method based on improved deep multi-sphere support vector data description 基于改进的深度多球支持向量数据描述的电梯门异常检测方法

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-28 DOI: 10.1016/j.compeleceng.2024.109660

Pengdong Xie , Linxuan Zhang , Minghong Li , Chaojie Qiu

Various types of elevator door faults and difficulties in fault data acquisition make it difficult to use supervised learning methods for fault diagnosis. This paper proposes a semi-supervised anomaly detection method based on improved deep multi-sphere support vector data description. Multiple distinguishing hyper-spheres, characterized by minimum volume, are established on the foundation of normal data by this method. These hyper-spheres represent the multi-modal distribution exhibited by the normal data. In addition, the method fuses multi-sensor source data such as tri-axial acceleration, dual-axial tilt angle, and introduces the structure of InceptionTime to realize the fusion of multivariate data and feature extraction in multiple resolutions. Experiments verify the feasibility of the method with an overall AUC of 96.50%, and comparative experiments demonstrate the superior detection performance. This contributes a novel, accurate, and more appropriate method to the elevator door anomaly detection.

电梯门故障种类繁多，故障数据获取困难，因此很难使用监督学习方法进行故障诊断。本文提出了一种基于改进的深度多球支持向量数据描述的半监督异常检测方法。该方法在正常数据的基础上建立了以最小体积为特征的多个区分超球。这些超球代表了正常数据表现出的多模态分布。此外，该方法还融合了三轴加速度、双轴倾斜角等多传感器源数据，并引入了 InceptionTime 结构，以实现多元数据的融合和多分辨率的特征提取。实验验证了该方法的可行性，总体 AUC 为 96.50%，对比实验也证明了其卓越的检测性能。这为电梯门异常检测提供了一种新颖、准确和更合适的方法。

引用次数: 0

A robust solution for recognizing accurate handwritten text extraction using quantum convolutional neural network and transformer models 利用量子卷积神经网络和变换器模型识别准确手写文本提取的稳健解决方案

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-28 DOI: 10.1016/j.compeleceng.2024.109794

Chiguru Aparna, K Rajchandar

Handwritten text extraction from images is challenging due to the variability of styles in handwriting, quality of images, and noise backgrounds. Existing methods often struggle to achieve high accuracy, hindering document analysis, optical character recognition, and data entry applications. We propose a novel approach to improve extraction accuracy, combining Quantum convolutional neural networks (QCNN) and transformer-based neural networks (TextExtractNet) named QTEN. Our method leverages the strengths of both models to recognize and extract handwritten text from images. Experimental results show that our approach achieves a 96 % accuracy rate, outperforming existing methods. This breakthrough has significant implications for automating document processing, data entry, and related applications. Our method's robustness and accuracy make it a valuable tool for industries relying on handwritten document processing, such as healthcare, finance, and government.

从图像中提取手写文本具有挑战性，这是因为手写体风格多变、图像质量和背景噪音等原因。现有的方法往往难以达到较高的准确率，从而阻碍了文档分析、光学字符识别和数据录入应用。我们提出了一种提高提取准确性的新方法，将量子卷积神经网络（QCNN）和基于变压器的神经网络（TextExtractNet）结合起来，命名为 QTEN。我们的方法充分利用了这两种模型的优势，从图像中识别并提取手写文本。实验结果表明，我们的方法达到了 96% 的准确率，优于现有方法。这一突破对自动化文档处理、数据录入和相关应用具有重要意义。我们的方法的鲁棒性和准确性使其成为医疗保健、金融和政府等依赖手写文档处理的行业的重要工具。

引用次数: 0

Multi-objective optimization in fixed-outline floorplanning with reinforcement learning 利用强化学习进行固定外线平面规划中的多目标优化

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-28 DOI: 10.1016/j.compeleceng.2024.109784

Zhongjie Jiang , Zhiqiang Li , Zhenjie Yao

Floorplanning is a crucial step in integrated circuit design. To address the fixed-outline floorplanning problem more effectively, we formulate it as a multi-objective optimization issue and employ multi-objective simulated annealing to simultaneously optimize both area and wirelength. Additionally, we apply deep reinforcement learning to learn from optimization experiences. This enables the exploration of more balanced multi-objective heuristics, thereby improving the results of multi-objective optimization. Test results on public benchmarks demonstrate the robust generalization capabilities of the proposed model. Compared to other advanced methods, our approach not only ensures a 100% success rate but also delivers superior performance in terms of wirelength. The deep reinforcement learning-assisted multi-objective simulated annealing method proposed in this paper can effectively address the fixed-outline floorplanning problem.

平面规划是集成电路设计的关键步骤。为了更有效地解决固定出线平面规划问题，我们将其表述为一个多目标优化问题，并采用多目标模拟退火同时优化面积和线长。此外，我们还应用了深度强化学习来学习优化经验。这样就能探索出更平衡的多目标启发式方法，从而改善多目标优化的结果。在公共基准上的测试结果表明，所提出的模型具有强大的泛化能力。与其他先进方法相比，我们的方法不仅确保了 100% 的成功率，而且在线长方面也表现出色。本文提出的深度强化学习辅助多目标模拟退火方法能有效解决固定外线楼层规划问题。

引用次数: 0

MonoCAPE: Monocular 3D object detection with coordinate-aware position embeddings MonoCAPE：利用坐标感知位置嵌入进行单目三维物体检测

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-25 DOI: 10.1016/j.compeleceng.2024.109781

Wenyu Chen , Mu Chen , Jian Fang , Huaici Zhao , Guogang Wang

3D monocular detection remains to be a focal point of research, particularly due to its capacity to deliver available precision under conditions of low cost and simplified configurations, making it especially valuable in fields like autonomous driving. Current 3D object detection methods often overlook the spatial information missing from images, which is critical to spatial perception, and optimize bounding box attributes separately, failing to meet the requirements of autonomous driving. We introduce MonoCAPE, a novel 3D detection framework addressing these issues by encoding spatial information and co-optimizing attributes through a Coordinate-Aware Position Encoding (CAPE) Generator and a Task Co-optimization Strategy (TCS). The CAPE Generator produces sparse positional embeddings, enabling spatial awareness with low computational cost, while the TCS utilizes Gaussian modeling to prevent suboptimal outputs. In this way, our framework comprehensively takes into account what existing approaches ignore. Extensive experiments on the KITTI dataset demonstrate MonoCAPE significantly improves

A P_{3 D}

and

A P_{B E V}

metrics compared to existing advanced methods.

三维单目检测仍然是研究的焦点，特别是因为它能够在低成本和简化配置的条件下提供可用精度，这使其在自动驾驶等领域尤为重要。目前的三维物体检测方法往往忽略了图像中缺失的空间信息，而这些信息对空间感知至关重要，并且单独优化边界框属性，无法满足自动驾驶的要求。我们介绍的 MonoCAPE 是一种新型三维检测框架，它通过坐标感知位置编码（CAPE）生成器和任务协同优化策略（TCS）对空间信息进行编码并对属性进行协同优化，从而解决这些问题。CAPE 生成器可生成稀疏的位置嵌入，从而以较低的计算成本实现空间感知，而 TCS 则利用高斯建模来防止次优输出。通过这种方式，我们的框架全面考虑了现有方法所忽略的问题。在 KITTI 数据集上进行的大量实验表明，与现有的先进方法相比，MonoCAPE 能显著提高 AP3D 和 APBEV 指标。

{"title":"MonoCAPE: Monocular 3D object detection with coordinate-aware position embeddings","authors":"Wenyu Chen , Mu Chen , Jian Fang , Huaici Zhao , Guogang Wang","doi":"10.1016/j.compeleceng.2024.109781","DOIUrl":"10.1016/j.compeleceng.2024.109781","url":null,"abstract":"<div><div>3D monocular detection remains to be a focal point of research, particularly due to its capacity to deliver available precision under conditions of low cost and simplified configurations, making it especially valuable in fields like autonomous driving. Current 3D object detection methods often overlook the spatial information missing from images, which is critical to spatial perception, and optimize bounding box attributes separately, failing to meet the requirements of autonomous driving. We introduce MonoCAPE, a novel 3D detection framework addressing these issues by encoding spatial information and co-optimizing attributes through a Coordinate-Aware Position Encoding (CAPE) Generator and a Task Co-optimization Strategy (TCS). The CAPE Generator produces sparse positional embeddings, enabling spatial awareness with low computational cost, while the TCS utilizes Gaussian modeling to prevent suboptimal outputs. In this way, our framework comprehensively takes into account what existing approaches ignore. Extensive experiments on the KITTI dataset demonstrate MonoCAPE significantly improves <span><math><mrow><mi>A</mi><msub><mrow><mi>P</mi></mrow><mrow><mn>3</mn><mi>D</mi></mrow></msub></mrow></math></span> and <span><math><mrow><mi>A</mi><msub><mrow><mi>P</mi></mrow><mrow><mi>B</mi><mi>E</mi><mi>V</mi></mrow></msub></mrow></math></span> metrics compared to existing advanced methods.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"120 ","pages":"Article 109781"},"PeriodicalIF":4.0,"publicationDate":"2024-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142533701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Synthetic aperture image enhancement with near-coinciding Nonuniform sampling case 合成孔径图像增强与近乎吻合非均匀采样情况

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-25 DOI: 10.1016/j.compeleceng.2024.109818

Xuebo Zhang , Peixuan Yang , Dengyu Cao

Multireceiver synthetic aperture sonar (SAS) uses multiple receivers to collect echoed signals, leading to nonuniform sampling in the azimuth dimension if the moving distance between adjacent pings is not half the receiver array length. The filter bank reconstruction (FBR) method, based on matrix inversion, is commonly used to address this by reconstructing uniform data from nonuniform signals. However, in practice, near-coinciding sampling can occur due to inaccuracies in the towed velocity of the SAS system, influenced by ocean conditions. The signal correlation of these near-coinciding receivers is very strong, and thus the steering vectors corresponding to these receivers are not completely independent. This further leads to an ill-conditioned system transfer function matrix made up of the receiver steering vectors, and results in inaccurate calculations of inverse matrix. Consequently, the FBR method suffers from significant signal-to-noise ratio loss or fails to reconstruct uniform signals accurately. This degradation in signal quality directly impacts the accuracy of target reconstruction, leading to errors in identifying and locating targets. This paper quantitatively defines nonuniform sampling with near-coinciding samples and discusses performance loss in various cases. To enhance imaging performance, we propose a method for reconstructing uniform signals. Simulation results demonstrate that the proposed method outperforms the conventional FBR method, providing better target reconstruction and higher efficiency.

多接收器合成孔径声纳（SAS）使用多个接收器收集回波信号，如果相邻 pings 之间的移动距离不是接收器阵列长度的一半，就会导致方位维度的不均匀采样。基于矩阵反演的滤波器组重建（FBR）方法通常用于解决这一问题，即从非均匀信号中重建均匀数据。然而，在实际应用中，由于 SAS 系统的拖曳速度受海洋条件影响而存在误差，可能会出现近重合采样的情况。这些近重合接收器的信号相关性非常强，因此这些接收器对应的转向矢量并非完全独立。这进一步导致由接收器转向矢量组成的系统传递函数矩阵条件不佳，并导致逆矩阵计算不准确。因此，FBR 方法会出现明显的信噪比损失，或无法准确重建均匀信号。信号质量的下降直接影响目标重建的准确性，导致目标识别和定位错误。本文定量定义了近重合采样的非均匀采样，并讨论了各种情况下的性能损失。为了提高成像性能，我们提出了一种重建均匀信号的方法。仿真结果表明，所提出的方法优于传统的 FBR 方法，能提供更好的目标重建和更高的效率。

{"title":"Synthetic aperture image enhancement with near-coinciding Nonuniform sampling case","authors":"Xuebo Zhang , Peixuan Yang , Dengyu Cao","doi":"10.1016/j.compeleceng.2024.109818","DOIUrl":"10.1016/j.compeleceng.2024.109818","url":null,"abstract":"<div><div>Multireceiver synthetic aperture sonar (SAS) uses multiple receivers to collect echoed signals, leading to nonuniform sampling in the azimuth dimension if the moving distance between adjacent pings is not half the receiver array length. The filter bank reconstruction (FBR) method, based on matrix inversion, is commonly used to address this by reconstructing uniform data from nonuniform signals. However, in practice, near-coinciding sampling can occur due to inaccuracies in the towed velocity of the SAS system, influenced by ocean conditions. The signal correlation of these near-coinciding receivers is very strong, and thus the steering vectors corresponding to these receivers are not completely independent. This further leads to an ill-conditioned system transfer function matrix made up of the receiver steering vectors, and results in inaccurate calculations of inverse matrix. Consequently, the FBR method suffers from significant signal-to-noise ratio loss or fails to reconstruct uniform signals accurately. This degradation in signal quality directly impacts the accuracy of target reconstruction, leading to errors in identifying and locating targets. This paper quantitatively defines nonuniform sampling with near-coinciding samples and discusses performance loss in various cases. To enhance imaging performance, we propose a method for reconstructing uniform signals. Simulation results demonstrate that the proposed method outperforms the conventional FBR method, providing better target reconstruction and higher efficiency.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"120 ","pages":"Article 109818"},"PeriodicalIF":4.0,"publicationDate":"2024-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142526693","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Offshore wind farms interfacing using HVAC-HVDC schemes: A review 使用 HVAC-HVDC 方案连接海上风电场：综述

IF 4 3区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computers & Electrical Engineering

Pub Date : 2024-10-25 DOI: 10.1016/j.compeleceng.2024.109797

Chen Zhichu , Mohsin Ali Koondhar , Ghulam Sarwar Kaloi , Muhammad Zain Yousaf , Aamir Ali , Zuhair Muhammed Alaas , Belgacem Bouallegue , Abdelmoty M. Ahmed , Yasser Ahmed Elshrief

Offshore wind farms (OWF) have emerged as a pivotal component in the transition towards renewable energy, offering substantial potential for reducing carbon emissions and enhancing energy security. Integration these offshore installations into the existing power grid present opportunities and challenges, particularly in terms of efficiency, stability, and cost-effectiveness. This review explores the interfacing of OWF with the power grid through High Voltage Alternating Current (HVAC) and High Voltage Direct Current (HVDC) schemes. This research offers a thorough examination of the latest technologies, and comparing the benefits and drawbacks of HVAC and HVDC systems for offshore wind applications. The review delves into various aspects, including technological advancements, economic considerations, and environmental impacts of HVAC and HVDC technologies. The review also addresses recent developments and contrasting viewpoints in the field, highlighting innovations and ongoing debates related to integrating OWF. By synthesizing recent studies and industry reports, this paper offers a balanced overview of the strengths and weaknesses associated with HVAC and HVDC schemes. It identifies critical areas for future research and provides recommendations for optimizing offshore wind farm integration. This review seeks to enhance understanding of current technologies and support decision-making in the development and deployment of offshore wind energy infrastructure.

海上风电场（OWF）已成为向可再生能源过渡的关键组成部分，为减少碳排放和提高能源安全提供了巨大潜力。将这些海上设施与现有电网整合既是机遇也是挑战，尤其是在效率、稳定性和成本效益方面。本综述探讨了通过高压交流电 (HVAC) 和高压直流电 (HVDC) 方案将海上风电场与电网连接的问题。本研究对最新技术进行了深入探讨，并比较了 HVAC 和 HVDC 系统在海上风电应用中的优缺点。综述深入探讨了各个方面，包括 HVAC 和 HVDC 技术的技术进步、经济因素和环境影响。综述还探讨了该领域的最新发展和不同观点，重点介绍了与集成海上风电场有关的创新和正在进行的讨论。通过综合最近的研究和行业报告，本文对暖通空调和高压直流方案的优缺点进行了均衡的概述。它确定了未来研究的关键领域，并提出了优化海上风电场集成的建议。本综述旨在加深对当前技术的理解，并为海上风能基础设施的开发和部署提供决策支持。

{"title":"Offshore wind farms interfacing using HVAC-HVDC schemes: A review","authors":"Chen Zhichu , Mohsin Ali Koondhar , Ghulam Sarwar Kaloi , Muhammad Zain Yousaf , Aamir Ali , Zuhair Muhammed Alaas , Belgacem Bouallegue , Abdelmoty M. Ahmed , Yasser Ahmed Elshrief","doi":"10.1016/j.compeleceng.2024.109797","DOIUrl":"10.1016/j.compeleceng.2024.109797","url":null,"abstract":"<div><div>Offshore wind farms (OWF) have emerged as a pivotal component in the transition towards renewable energy, offering substantial potential for reducing carbon emissions and enhancing energy security. Integration these offshore installations into the existing power grid present opportunities and challenges, particularly in terms of efficiency, stability, and cost-effectiveness. This review explores the interfacing of OWF with the power grid through High Voltage Alternating Current (HVAC) and High Voltage Direct Current (HVDC) schemes. This research offers a thorough examination of the latest technologies, and comparing the benefits and drawbacks of HVAC and HVDC systems for offshore wind applications. The review delves into various aspects, including technological advancements, economic considerations, and environmental impacts of HVAC and HVDC technologies. The review also addresses recent developments and contrasting viewpoints in the field, highlighting innovations and ongoing debates related to integrating OWF. By synthesizing recent studies and industry reports, this paper offers a balanced overview of the strengths and weaknesses associated with HVAC and HVDC schemes. It identifies critical areas for future research and provides recommendations for optimizing offshore wind farm integration. This review seeks to enhance understanding of current technologies and support decision-making in the development and deployment of offshore wind energy infrastructure.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"120 ","pages":"Article 109797"},"PeriodicalIF":4.0,"publicationDate":"2024-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142533715","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0