检索结果-内蒙古大学图书馆

IEEE Transactions on Circuits and Systems for Artificial Intelligence 2025年第2期2卷 114-125页

作者： Mikail Yayla Ming-Liang Wei Simon Thomann Albi Mema Chia-Lin Yang Jian-Jia Chen Hussam Amrouch Design Automation for Embedded Systems Group TU Dortmund University Dortmund Germany Lamarr Institute for Machine Learning and Artificial Intelligence Dortmund Germany Department of the Graduate Institute of Electronics Engineering National Taiwan University Taipei Taiwan Technical University of Munich (TUM) Munich Germany TUM School of Computation Information and Technology Chair of AI Processor Design Munich Institute of Robotics and Machine Intelligence Munich Germany Department of Computer Science and Information Engineering Graduate Institute of Electronics Engineering National Taiwan University Taipei Taiwan Department of Computer Science TU Dortmund University Dortmund Germany Lamarr Institute Dortmund Germany

Accelerating the immense workloads of Neural Networks (NNs) is a critical challenge, and analog computing-based approaches present a promising solution. Among these, the Integrate-and-Fire (IF) Spiking Neural Network (SNN) stands out for its potential in efficient neural computation. However, achieving high inference accuracy through precise multiply-accumulate (MAC) operations necessitates a large membrane capacitor—224 to 547 times larger than the synapse array at the 14nm technology node—resulting in prohibitive area costs. Additionally, the large capacitor size introduces higher energy consumption and longer access times. Thus, a key research challenge is maintaining accuracy while mitigating the costs associated with the large membrane capacitor. In this work, we propose a HW/SW Codesign method, called CapMin, for capacitor size minimization in analog computing IF-SNNs. CapMin minimizes the capacitor size by reducing the number of spike times needed for accurate operation of the HW, based on the absolute frequency of MAC level occurrences in the SW. To increase the operation of IF-SNNs to current variation, we propose the method CapMin-V, which trades capacitor size for protection based on the reduced capacitor size found in CapMin. In our experiments, CapMin achieves more than a 14 $\times$ reduction in capacitor size along with a 34% reduction in energy consumption over the state of the art, while CapMin-V achieves increased variation tolerance in the IF-SNN operation, requiring only a small increase in capacitor size.

关键词： Capacitors Circuits Accuracy Capacitance Micromechanical devices Costs Neurons Integrated circuit modeling Spiking neural networks Hardware

来源：评论

学校读者我要写书评

暂无评论

Perceptive self-supervised learning network for noisy image watermark removal

arXiv

引用

arXiv 2024年

作者： Tian, Chunwei Zheng, Menghua Li, Bo Zhang, Yanning Zhang, Shichao Zhang, David School of Software Northwestern Polytechnical University Xi’an710129 China National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Xi’an710129 China School of Electronics and Information Northwestern Polytechnical University Xi’an710129 China School of Computer Science Northwestern Polytechnical University National Engineering Laboratory for Integrated AeroSpace-Ground-Ocean Big Data Application Technology Xi’an710129 China Guangxi Key Lab of Multisource Information Mining & Security College of Computer Science & Engineering Guangxi Normal University Guilin541004 China Shenzhen518172 China Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen518172 China

Popular methods usually use a degradation model in a supervised way to learn a watermark removal model. However, it is true that reference images are difficult to obtain in the real world, as well as collected images by cameras suffer from noise. To overcome these drawbacks, we propose a perceptive self-supervised learning network for noisy image watermark removal (PSLNet) in this paper. PSLNet depends on a parallel network to remove noise and watermarks. The upper network uses task decomposition ideas to remove noise and watermarks in sequence. The lower network utilizes the degradation model idea to simultaneously remove noise and watermarks. Specifically, mentioned paired watermark images are obtained in a self-supervised way, and paired noisy images (i.e., noisy and reference images) are obtained in a supervised way. To enhance the clarity of obtained images, interacting two sub-networks and fusing obtained clean images are used to improve the effects of image watermark removal in terms of structural information and pixel enhancement. Taking into texture information account, a mixed loss uses obtained images and features to achieve a robust model of noisy image watermark removal. Comprehensive experiments show that our proposed method is very effective in comparison with popular convolutional neural networks (CNNs) for noisy image watermark removal. Codes can be obtained at https://***/hellloxiaotian/PSLNet. Copyright © 2024, The Authors. All rights reserved.

关键词： Image denoising

来源：评论

学校读者我要写书评

暂无评论

Arthroscopic multi-spectral scene segmentation using deep learning

arXiv

引用

arXiv 2021年

作者： Ali, Shahnewaz Jonmohamadi, Yaqub Takeda, Yu Roberts, Jonathan Crawford, Ross Brown, Cameron Pandey, Ajay K. Robotics and Autonomous Systems School of Electrical Engineering and Computer Science Queensland University of Technology Gardens Point BrisbaneQLD4001 Australia

Knee arthroscopy is a minimally invasive surgical (MIS) procedure which is performed to treat knee-joint ailment. Lack of visual information of the surgical site obtained from miniaturized cameras make this surgical procedure more complex. Knee cavity is a very confined space;therefore, surgical scenes are captured at close proximity. Insignificant context of knee atlas often makes them unrecognizable as a consequence unintentional tissue damage often occurred and shows a long learning curve to train new surgeons. Automatic context awareness through labeling of the surgical site can be an alternative to mitigate these drawbacks. However, from the previous studies, it is confirmed that the surgical site exhibits several limitations, among others, lack of discriminative contextual information such as texture and features which drastically limits this vision task. Additionally, poor imaging conditions and lack of accurate ground-truth labels are also limiting the accuracy. To mitigate these limitations of knee arthroscopy, in this work we proposed a scene segmentation method that successfully segments multi structures. Copyright © 2021, The Authors. All rights reserved.

关键词： Spectroscopy

来源：评论

学校读者我要写书评

暂无评论

Navigate-and-seek: A robotics framework for people localization in agricultural environments

arXiv

引用

arXiv 2021年

作者： Polvara, Riccardo Del Duchetto, Francesco Neumann, Gerhard Hanheide, Marc School of Computer Science University of Lincoln United Kingdom Saga Robotics Lincoln United Kingdom Karlsruhe Institute of Technology Karlsruhe Germany

The agricultural domain offers a working environment where many human laborers are nowadays employed to maintain or harvest crops, with huge potential for productivity gains through the introduction of robotic automation. Detecting and localizing humans reliably and accurately in such an environment, however, is a prerequisite to many services offered by fleets of mobile robots collaborating with human workers. Consequently, in this paper, we expand on the concept of a topological particle filter (TPF) to accurately and individually localize and track workers in a farm environment, integrating information from heterogeneous sensors and combining local active sensing (exploiting a robot's onboard sensing employing a Next-Best-Sense planning approach) and global localization (using affordable IoT GNSS devices). We validate the proposed approach in topologies created for the deployment of robotics fleets to support fruit pickers in a real farm environment. By combining multi-sensor observations on the topological level complemented by active perception through the NBS approach, we show that we can improve the accuracy of picker localization in comparison to prior work. Copyright © 2021, The Authors. All rights reserved.

关键词： Productivity

来源：评论

学校读者我要写书评

暂无评论

New High Boost DC-DC Converter for Renewable Energy Conversion

New High Boost DC-DC Converter for Renewable Energy Conversi...

引用

International Conference on Communications and Electronics, ICCE

作者： Dinh-Tuyen Nguyen Truong-Khang Duong Truong-Duy Duong Trong-Nha Quang Ton Duc Do Faculty of Electrical and Electronics Engineering Ho Chi Minh City University of TechnologyVNU-HCM Ho Chi Minh City Vietnam Dept. Electrical and Computer Engineering Wayne State University Detroit MI USA High Volume Power Design Advanced Energy Industries Inc Eden Prairie MN USA Dept. Robotics and Mechatronics School of Science and Technology Nazarbayev University Kazakhstan

ISBN: (数字)9798350379792

ISBN: (纸本)9798350379808

This paper proposes a high-boost power converter for renewable energy systems. The suggested converter exhibits a simplified structure by replacing a diode in the switched-inductor cell with an active power switch to reduce the total voltage stress on the semiconductor device, get cost reduction, and improve efficiency. Additionally, the common ground feature inherent in the introduced topology expands its potential for renewable energy applications. Finally, the simulation verifications are performed using PSIM software to confirm the effectiveness of the suggested approach, and a detailed comparison study between the basic switched-inductor converter and the suggested converter is also presented.

关键词： Renewable energy sources Costs Simulation Prototypes Voltage Switches Software

来源：评论

学校读者我要写书评

暂无评论

Boosting the Transferability of Adversarial Attacks with Global Momentum Initialization

arXiv

引用

arXiv 2022年

作者： Wang, Jiafeng Chen, Zhaoyu Jiang, Kaixun Yang, Dingkang Hong, Lingyi Guo, Pinxue Guo, Haijing Zhang, Wenqiang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai200433 China Shanghai Engineering Research Center of AI & Robotics Academy for Engineering & Technology Fudan University Shanghai200433 China Engineering Research Center of Robotics Ministry of Education Academy for Engineering & Technology Fudan University Shanghai200433 China

Deep Neural Networks (DNNs) are vulnerable to adversarial examples, which are crafted by adding human-imperceptible perturbations to the benign inputs. Simultaneously, adversarial examples exhibit transferability across models, enabling practical black-box attacks. However, existing methods are still incapable of achieving the desired transfer attack performance. In this work, focusing on gradient optimization and consistency, we analyse the gradient elimination phenomenon as well as the local momentum optimum dilemma. To tackle these challenges, we introduce Global Momentum Initialization (GI), providing global momentum knowledge to mitigate gradient elimination. Specifically, we perform gradient pre-convergence before the attack and a global search during this stage. GI seamlessly integrates with existing transfer methods, significantly improving the success rate of transfer attacks by an average of 6.4% under various advanced defense mechanisms compared to the state-of-the-art method. Ultimately, GI demonstrates strong transferability in both image and video attack domains. Particularly, when attacking advanced defense methods in the image domain, it achieves an average attack success rate of 95.4%. The code is available at https://***/Omenzychen/Global-Momentum-Initialization. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Study on spiral winding swimming motion control of a slender legless creature model

引用

High technology Letters 2021年第4期27卷 415-421页

作者： Lu Zhenli Ma Zhipeng Marko Pencic Maja Cavic Borovac Branislav Roumiana Ilieva Bojan Nemec Marjan Mernik School of Electrical Engineering and Automation Changshu Institute of TechnologyChangshu 215500P.R.China School of Information and Control Engineering China University of Mining and TechnologyXuzhou 221116P.R.China Faculty of Technical Sciences University of Novi SadNovi Sad 621000Serbia Faculty of Management Technical University of SofiaSofia 1000Bulgaria Department of Automatics Biocybernetics and RoboticsJožef Stefan InstituteLjubljana 1000Slovenia Faculty of Electrical Engineering and Computer Science University of MariborMaribor 462000Slovenia

Through the observation and analysis of the motion trajectory of spiral winding motion for slender legless biological creatures in water,V-REP software is adopted to build a dynamic simulator to study on the mechanism of spiral winding swimming(SWS)*** using the output of spiral function(SF),the dynamic simulation model of slender legless creature(SLC)realizes the SWS motion in *** corresponding experiments under the control of different bending angle of SF to control the dynamic model are also carried out to analyze the water performance of SWS in still *** with the output of two different bending angles of the spiral function,the dynamic model can be used to realize SWS and up/down *** work provides technical reserve and experimental platform for the corresponding study in related fields.

关键词： V-REP spiral winding swimming(SWS) motion control slender legless creature(SLC)

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of Nonlinear Effects in Frequency-Splitting SWIPT Signals 91

Performance Evaluation of Nonlinear Effects in Frequency-Spl...

引用

91st IEEE Vehicular technology Conference, VTC Spring 2020

作者： Rajaram, Akashkumar Dinis, Rui Madeira, Joao Jayakody, Dushantha Nalin K. Beko, Marko New University of Lisbon Faculty of Science and Technology Portugal National Research Tomsk Polytechnic University School of Computer Science and Robotics Russia Universidade Lusòfona Copelabs Lisbon Portugal

ISBN: (纸本)9781728152073

Simultaneous wireless information and power transmission (SWIPT) is one of the popular technique used in radio frequency energy harvesting (EH). Frequency splitting based SWIPT (FS-SWIPT) is a novel SWIPT technique that can be adopted with frequency domain multiple access signal transmission protocol for the near field EH application in biomedical sensors. In this paper, we implement FS-SWIPT in orthogonal frequency division multiplexed (OFDM) signal, the OFDM signal has separate sub carriers for the data symbols and the energy carrying symbols. At the receiver, the energy and information symbols are based on their sub carriers by using FS-SWIPT technique. The amplification of high energy OFDM signal at transmitter by using a high power amplifier are susceptible to non linear distortion (NLD), which degrades the bit error rate (BER) performance. We analyze the impact of NLD on the BER performance of FS-SWIPT and validate the analytical results by using simulations. © 2020 IEEE.

关键词： Energy harvesting

来源：评论

学校读者我要写书评

暂无评论

Extracting Semantic Knowledge from GANs with Unsupervised Learning

arXiv

引用

arXiv 2022年

作者： Xu, Jianjin Zhang, Zhaoxiang Hu, Xiaolin The School of Mathematics and Computer Science Panzhihua University The Department of Computer Science and Technology Tsinghua University China The Insitute of Automation Chinese Academy of Sciences The Centre for Artificial Intelligence and Robotics Hong Kong Innovation and Science Insitute Chinese Academy of Sciences Hong Kong The Department of Computer Science and Technology Institute for Artificial Intelligence State Key Laboratory of Intelligent Technology and Systems THU-Bosch JCML Center THBI IDG/McGovern Institute for Brain Research BNRist Tsinghua University Beijing China Beijing China

Recently, unsupervised learning has made impressive progress on various tasks. Despite the dominance of discriminative models, increasing attention is drawn to representations learned by generative models and in particular, Generative Adversarial Networks (GANs). Previous works on the interpretation of GANs reveal that GANs encode semantics in feature maps in a linearly separable form. In this work, we further find that GAN’s features can be well clustered with the linear separability assumption. We propose a novel clustering algorithm, named KLiSH, which leverages the linear separability to cluster GAN’s features. KLiSH succeeds in extracting fine-grained semantics of GANs trained on datasets of various objects, e.g., car, portrait, animals, and so on. With KLiSH, we can sample images from GANs along with their segmentation masks and synthesize paired image-segmentation datasets. Using the synthesized datasets, we enable two downstream applications. First, we train semantic segmentation networks on these datasets and test them on real images, realizing unsupervised semantic segmentation. Second, we train image-to-image translation networks on the synthesized datasets, enabling semantic-conditional image synthesis without human annotations. © 2022, CC BY-SA.

关键词： Unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Toward a Framework for Adaptive Impedance Control of an Upper-limb Prosthesis

arXiv

引用

arXiv 2022年

作者： Ferrante, Laura Sridharan, Mohan Zito, Claudio Farina, Dario School of Computer Science University of Birmingham Birmingham United Kingdom Autonomous Robotics Research Centre Technology Innovation Institute Abu Dhabi United Arab Emirates Department of Bioengineering Imperial College London London United Kingdom

Adapting upper-limb impedance (i.e., stiffness, damping, inertia) is essential for humans interacting with dynamic environments for executing grasping or manipulation tasks. On the other hand, control methods designed for state-of-the-art upper-limb prostheses infer motor intent from surface electromyography (sEMG) signals in terms of joint kinematics, but they fail to infer and use the underlying impedance properties of the limb. We present a framework that allows a human user to simultaneously control the kinematics, stiffness, and damping of a simulated robot through wrist’s flexion-extension. The framework includes muscle-tendon units and a forward dynamics block to estimate the motor intent from sEMG signals, and a variable impedance controller that implements the estimated intent on the robot, allowing the user to adapt the robot’s kinematics and dynamics online. We evaluate our framework with 8 able-bodied subjects and an amputee during reaching tasks performed in free space, and in the presence of unexpected external perturbations that require adaptation of the wrist impedance to ensure stable interaction with the environment. We experimentally demonstrate that our approach outperforms a data-driven baseline in terms of its ability to adapt to external perturbations, overall controllability, and feedback from participants. © 2022, CC BY.

关键词： Robots

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：