检索结果-内蒙古大学图书馆

5th International Conference on Computer, Big Data and Artificial intelligence, ICCBD+AI 2024

作者： Luo, Yujia Hu, Qingsong Cheng, Yuanxun Yuan, Shuya Zhang, Yuansheng China University of Mining and Technology School of Information and Control Engineering Xuzhou China State Key Laboratory of Intelligent Optimized Manufacturing in Mining & Metallurgy Process Beijing China Beijing Key Laboratory of Process Automation in Mining & Metallurgy Beijing China

ISBN: (纸本)9798331533991

With the advent of the 6G era, precise positioning and reliable, energy-efficient data transmission have become critical for the Internet of Things. However, high-frequency signals like millimeter-wave and terahertz are easily obstructed, making traditional LOS-based methods ineffective in complex environments. To overcome this, NLOS localization methods and Reconfigurable Intelligent Surfaces offer promising solutions by dynamically modifying the wireless environment. This paper introduces a Direct Position Determination method that integrates Deep Neural Networks (DNN) with the Sparrow Search Algorithm (SSA) to tackle the challenges of high computational complexity and low localization accuracy in RIS-aided systems. An NLOS scenario with RIS is established, and a far-field channel model is derived. The method uses DNN to model the nonlinear relationship between the likelihood function and user position, and then employs SSA to optimize the DNN-based fitness function, efficiently solving the objective function. Experimental results show that the DNN-SSA method consistently achieves high localization accuracy across different scenarios, even at low SNR levels, demonstrating its effectiveness and potential for 6G high-precision localization. © 2024 ieee.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Super Learner Ensemble for Anomaly Detection and Cyber-Risk Quantification in Industrial control Systems

引用

ieee INTERNET OF THINGS JOURNAL 2022年第15期9卷 13279-13297页

作者： Ahmadi-Assalemi, Gabriela Al-Khateeb, Haider Epiphaniou, Gregory Aggoun, Amar Univ Wolverhampton Sch Engn Comp & Math Sci Wolverhampton WV1 1LY England Univ Warwick Int Mfg Ctr Warwick Mfg Grp Coventry CV4 7AL W Midlands England

Industrial control systems (ICSs) are integral parts of smart cities and critical to modern societies. Despite indisputable opportunities introduced by disruptor technologies, they proliferate the cybersecurity threat landscape, which is increasingly more hostile. The quantum of sensors utilized by ICS aided by artificial intelligence (AI) enables data collection capabilities to facilitate automation, process streamlining, and cost reduction. However, apart from the operational use, the sensors generated data combined with AI can be innovatively utilized to model anomalous behavior as part of layered security to increase resilience to cyberattacks. We introduce a framework to profile anomalous behavior in ICS and derive a cyber-risk score. A novel super learner ensemble for one-class classification is developed, using overlapping rolling windows with stratified, k-fold, n-repeat cross-validation applied to each base learner followed by majority voting to derive the best learner. Our approach is demonstrated on a liquid distribution sensor data set. The experimental results reveal that the proposed technique achieves an overall F1-score of 99.13 %, an anomalous recall score of 99% detecting anomalies lasting only 17 s. The key strength of the framework is the low computational complexity and error rate. The framework is modular, generic, applicable to other ICS, and transferable to other smart city sectors.

关键词： Cyber resilience cyber security cyber-physical systems (CPSs) digital forensic and incident response (DFIR) human-machine interface (HMI) industry 4.0 insider threat Internet of Things (IoT) machine learning (ML) programmable logic controllers (PLC) smart city supervisory control and data acquisition (SCADA)

来源：评论

学校读者我要写书评

暂无评论

Self and Cross Motion Extracted in Drone Video to Improve Object Detection

Self and Cross Motion Extracted in Drone Video to Improve Ob...

引用

2023 ieee International Conference on Unmanned Systems, ICUS 2023

作者： Zhou, Zihao Yu, Xianguo Chen, Xiangcheng Li, Yuke School of Automation Wuhan University of Technology Wuhan China College of Intelligence Science and Technology National University of Defense Technology Changsha China School of Artificial Intelligence Anhui University Hefei China

ISBN: (纸本)9798350316308

The abundant temporal information presented in videos can effectively address the issue of significant appearance deterioration observed in drone images. Learning temporal features in videos poses a significant challenge. Simple feature aggregation methods often result in the interference of background information into targets. To address this issue, we propose a simple yet effective Self and Cross Motion Extracted (SACME). The key idea is to leverage transformer layers and a novel temporal fusion network to learn and fuse temporal features. SACME serves as an addon module, enabling existing static object detectors to achieve high-performance video object detection without incurring significant additional computational costs. To validate the efficacy of our module, we select YOLOv7 as baseline and carried out comprehensive experiments on the VisDrone2019-VID dataset. Particularly, our SACME-YOLOv7 not only achieves a significant 5.1 % improvement in mean average precision (mAP) on the challenging VisDrone dataset, but also achieves a remarkable running speed of 31 frames per second (fps). © 2023 ieee.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

Application of Fuzzy Theory to Parameter Design in Wafer Stealth Laser Dicing

Application of Fuzzy Theory to Parameter Design in Wafer Ste...

引用

2024 International symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2024

作者： Su, Te-Jen Chiu, Chien-Liang Su, Chun-Hsien Tian, Mau-Yi Cheng, Jui-Chuan Wang, Shih-Ming Department of Electronic Engineering National Kaohsiung University of Sciences and Technology Kaohsiung Taiwan Department of Computer Science and Information Engineering Cheng Shiu University Kaohsiung Taiwan

ISBN: (纸本)9798350389210

The application of fuzzy theory in stealth laser dicing technology for wafers aims to enhance the precision and efficiency of cutting parameters. With technological advancements and the growth of the electric vehicle industry, the demand for semiconductor chips has rapidly increased, driving wafer cutting technology towards greater efficiency and precision. Compared to traditional blade cutting and laser cutting techniques, stealth laser dicing technology can avoid the chipping caused by blade cuts and the thermal damage caused by laser cuts, providing a stress-free and thermal damage-free solution. This study optimizes cutting parameters through a fuzzy control system using a layered setting method. The first layer determines the most suitable laser power based on the thickness of the workpiece and the metal layer;the second layer uses the final wafer thickness and the laser power determined in the first layer as inputs to calculate the optimal scribing height. These parameter settings aim to enhance cutting quality and efficiency while also verifying the practical application effects of the fuzzy control system. Experimental results show that the fuzzy control system significantly improves cutting precision and efficiency, achieving a 100% cutting success rate, a 66% reduction in material consumption, a 4% reduction in energy usage, and a 50% reduction in parameter adjustment labor time. The precision of the fuzzy control system in controlling laser power surpasses manual settings, and the optimization of scribing height is also significantly improved. The study demonstrates that fuzzy theory has extensive potential applications in stealth laser dicing of wafers. Future research may integrate artificial intelligence technology to further enhance the automation of cutting parameter settings, thereby improving production efficiency and product quality. © 2024 ieee.

关键词： Fuzzy control

来源：评论

学校读者我要写书评

暂无评论

Research on human movements push up counter based on media-pipe artificial intelligence 6

Research on human movements push up counter based on media-p...

引用

6th Asia symposium on Image Processing, ASIP 2024

作者： Saido, Balde Lu, Guangda Qin, Zhuanpin Yin, Xianhan Dai, Jianyi Guo, Xiaoqian Tianjin University of Technology and Education Tianjin Key Laboratory of Information Sensing and Intelligent Control Hexi District No. 1310 Tianjin300222 China

ISBN: (纸本)9798350376159

This paper introduces a project that integrates pose estimation and deep learning to evaluate exercise performance and provide real-time feedback on repetition counts during workouts. Using machine learning in the fitness sector, the study focuses on accurately tracking and counting exercise repetitions, which is beneficial for physical education and competitive fitness events. The paper presents a novel real-time push-up counting method based on 2D video analysis, utilizing Media-pipe for precise pose estimation and motion analysis. By following the push-up guidelines set by the National Academy of Sports Medicine, the method differentiates between correct and incorrect push-up forms. The experimental results show that the proposed method achieves an accuracy of 90.00%, precision of 91.67%, recall of 91.67%, and an F1-Score of 91.66%. The integration of pose estimation and deep learning in the fitness industry offers new opportunities for improving workout monitoring, ensuring proper form, and optimizing fitness outcomes for individuals of different fitness levels. © 2024 ieee.

关键词： Sports medicine

来源：评论

学校读者我要写书评

暂无评论

Training and Analysis of Hyperparameters in Neural Networks for Computer Vision Applications: A Didactic Approach 20

Training and Analysis of Hyperparameters in Neural Networks ...

引用

20th ieee Jubilee World symposium on Applied Machine intelligence and Informatics, SAMI 2022

作者： Barreto, Saulo Cardoso Szemes, Peter Tamas Faculty of Engineering Department of Mechatronics Debrecen Hungary

ISBN: (纸本)9781665497046

Machine Learning models are known for having millions of parameters considering different applications, such as object recognition and self-driving cars. To avoid spending more time than necessary in the training stage and more importantly, to achieve a satisfactory model, the hyperparameters should be correctly defined. The main goal of the proposed project was to identify and evaluate the influence of hyperparameters in the training of neural networks for computer vision applications. Besides that, we aim to provide a didactic approach to analyze the acquired generalization capability of a model and easy visualization of the impact of different parameters during the training phase. © 2022 ieee.

关键词： Training Analytical models Computer vision computational modeling Neural networks Machine learning Data models

来源：评论

学校读者我要写书评

暂无评论

Logic-based machine comprehension for chatbots 20

Logic-based machine comprehension for chatbots

引用

20th ieee Jubilee World symposium on Applied Machine intelligence and Informatics, SAMI 2022

作者： Redeca, Sergiu Groza, Adrian Technical University of Cluj-Napoca Romania

ISBN: (纸本)9781665497046

We target here the interpretation of natural language by means of Equational First Order Logic (FOL). The communicative acts of the human agent are automatically converted into a First Order Logic theory. The models of the FOL theory are analysed by Mace4 model finder. To goal of the chatbot is to reduce the number of interpretation models. With a single interpretation model, the chatbot is sure about the human agent preferences. To reduce the number of models, the agent asks questions, while the corresponding answers are used to restrict the models. The order in which these questions are asked is extremely relevant to effectevely reduce the huge number of interpretation models of a FOL theory. We apply here an entropy-based method to improve the dialogues between a conversational agent and a human agen. The new method computes the entropy on the set of possible interpretation models. The experiments indicated that with entropy-based chatbots it is easier to estimate the budget of questions needed to elicit client preferences. © 2022 ieee.

关键词： Analytical models computational modeling Natural languages Oral communication Chatbots Entropy Mathematical models

来源：评论

学校读者我要写书评

暂无评论

Non-Prehensile Whole-Arm Manipulation Grasping and Trajectory Generation in 3D 3

Non-Prehensile Whole-Arm Manipulation Grasping and Trajector...

引用

3rd International Conference on Robotics, automation and Artificial intelligence (RAAI)

作者： Hinrichs, Pascal Kowalski, Christian Pfingsthorn, Max Hein, Andreas OFFIS Inst Informat Technol Oldenburg Germany Carl von Ossietzky Univ Oldenburg Div Assistance Syst & Med Device Technol Oldenburg Germany

ISBN: (纸本)9798350307573;9798350307566

For manipulating large, bulky loads, whole-arm manipulation is a suitable method. It is often used in a non-prehensile way. However, non-prehensile manipulation is mostly considered only two-dimensional, for example, moving an object on a surface. Whole-arm manipulation, on the other hand, often requires 3-dimensional movements. If redundant manipulators are used for whole-arm manipulation, there may be several valid poses for a way-point on a trajectory but only one of these configurations allows to cover the largest possible distance on the trajectory without reconfiguration. Reconfiguration is not possible with whole-arm manipulation, as this would result in a loss of the object. Numerical methods, on the contrary, can only randomly select one of these configurations. Therefore, we propose a completely analytical trajectory generation method for moving cylindrical bodies, which allows exploration of the entire path for optimal robot configuration and redundancy resolution. The contact plane along the longitudinal axis of the cylinder can be passed as an allowed interval and is used by the algorithm for optimization. In experiments, we compare our method against RRTConnect and the impact of the shiftability of the contact plane on the performance of our method. The results show that the proposed method finds up to two and a half times longer distances than RRTConnect on a path without reconfiguration, but takes more time to do so. Finally, we demonstrate the applicability in a real-world scenario in nursing.

关键词： Multi-contact whole-body Motion Planning and control Constrained Motion Planning Manipulation Planning Trajectory optimization computational geometry Manipulators Grasping

来源：评论

学校读者我要写书评

暂无评论

Data Augmentation for Vulnerability Detection Based on Code Refactoring and Mixup 35

Data Augmentation for Vulnerability Detection Based on Code ...

引用

35th ieee International symposium on Software Reliability Engineering Workshops, ISSREW 2024

作者： Xing, Ying Huang, Jiaqi Wang, Guilong You, Yehao Yang, Bin Li, Xiaofeng Luo, Yixing Beijing University of Posts and Telecommunications School of Intelligent Engineering and Automation Beijing China Beijing University of Posts and Telecommunications School of Artifcial Intelligence Beijing China Yunnan Key Laboratory of Software Engineering Yunnan China China Unicom Research Institute Beijing China Beijing Institute of Control Engineering Software Development Center Beijing China

ISBN: (纸本)9798350367041

In recent years, the number of software vulnerabilities has gradually increased, posing a threat to the security of software, and identifying whether there are vulnerabilities in software is crucial for assuring its quality. With the development of deep learning, vulnerability detection based on deep learning has achieved good results. However, real-world software vulnerability datasets often suffer from class imbalance and a lack of labeled data, limiting the performance of deep learning models. Therefore, this paper proposes a data augmentation method to alleviate these two problems. Firstly, we treats samples with vulnerabilities in the dataset as code, and generates more samples with vulnerabilities through code refactoring to alleviate class imbalance issues. Secondly, we use the mixup method to perform linear interpolation on the embedding vectors which represent codes, and add the interpolated vectors into the training set, thereby increasing the amount of labeled training data. To evaluate our method, we conducted experiments on three projects of the Promise dataset. The experimental results show that our data augmentation method can effectively improve the effect of vulnerability detection. © 2024 ieee.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Infrared video behavior recognition algorithm based on AGX Xavier environment

Infrared video behavior recognition algorithm based on AGX X...

引用

2023 CAA symposium on Fault Detection, Supervision and Safety for Technical Processes, SAFEPROCESS 2023

作者： Liu, Mingrui Wang, Xiaogang Wei, Yang Zhou, Jiayi Song, Rui Lian, Zewei Sichuan University of Science and Engineering School of Automation and Information Engineering Yibin644000 China Artificial Intelligence Key Laboratory of Sichuan Province Yibin China

ISBN: (纸本)9798350337754

Analyzing human behavior recognition for infrared video is important since crimes are more likely to occur in nighttime environments. Therefore, this paper proposes an infrared video behavior recognition algorithm based on AGX Xavier, an edge device, using a TSM module embedded in a 2D neural network with temporal modeling capability, which can significantly reduce the number of model parameters and computational effort to achieve deployment on a lightweight edge device with cameras. Online behavior recognition is achieved. The method's accuracy is 91% on the dataset captured by the subject's active infrared camera. It is tested in real application scenarios, and the online recognition mode has low latency and high speed to meet the real-time behavior recognition requirements. © 2023 ieee.

关键词： Cameras

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：