检索结果-内蒙古大学图书馆

Generalization-Enhanced Few-Shot Object Detection in Remote Sensing

IEEE Transactions on Circuits and Systems for Video technology 2025年第6期35卷 5445-5460页

作者： Lin, Hui Li, Nan Yao, Pengjuan Dong, Kexin Guo, Yuhan Hong, Danfeng Zhang, Ying Wen, Congcong China Academy of Electronics and Information Technology Beijing100846 China China Meteorological Administration National Satellite Meteorological Center Innovation Center for FengYun Meteorological Satellite Beijing100081 China Tsinghua University State Key Laboratory of Hydroscience and Engineering Department of Hydraulic Engineering Beijing100084 China Chinese Academy of Sciences Aerospace Information Research Institute Beijing100094 China University of Chinese Academy of Sciences School of Electronic Electrical and Communication Engineering Beijing100049 China University of Science and Technology Beijing School of Automation and Electrical Engineering Beijing100083 China New York University Abu Dhabi Department of Electrical and Computer Engineering Abu Dhabi United Arab Emirates University of Science and Technology of China Anhui China

Object detection is a fundamental task in computer vision that involves accurately locating and classifying objects within images or video frames. In remote sensing, this task is particularly challenging due to the high resolution, multi-scale features, and diverse ground object characteristics inherent in satellite and UAV imagery. These challenges necessitate more advanced approaches for effective object detection in such environments. While deep learning methods have achieved remarkable success in remote sensing object detection, they typically rely on large amounts of labeled data. Acquiring sufficient labeled data, particularly for novel or rare objects, is both challenging and time-consuming in remote sensing scenarios, limiting the generalization capabilities of existing models. To address these challenges, few-shot learning (FSL) has emerged as a promising approach, aiming to enable models to learn new classes from limited labeled examples. Building on this concept, few-shot object detection (FSOD) specifically targets object detection challenges in data-limited conditions. However, the generalization capability of FSOD models, particularly in remote sensing, is often constrained by the complex and diverse characteristics of the objects present in such environments. In this paper, we propose the Generalization-Enhanced Few-Shot Object Detection (GE-FSOD) model to improve the generalization capability in remote sensing FSOD tasks. Our model introduces three key innovations: the Cross-Level Fusion Pyramid Attention Network (CFPAN) for enhanced multi-scale feature representation, the Multi-Stage Refinement Region Proposal Network (MRRPN) for more accurate region proposals, and the Generalized Classification Loss (GCL) for improved classification performance in few-shot scenarios. GE-FSOD demonstrates superior robustness and accuracy in remote sensing FSOD tasks through these enhancements. Extensive experiments on the DIOR and NWPU VHR-10 datasets show that our

关键词： Labeled data

来源：评论

学校读者我要写书评

暂无评论

Characteristics, Classification and Challenges in Searching for CEMP Stars

引用

Research in Astronomy and Astrophysics 2025年第5期25卷 184-193页

作者： Lefeng He Xiangru Li Xiaoming Kong A-Li Luo Haifeng Yang School of Computer Science South China Normal University School of Mechanical Electrical&Information EngineeringShandong University Shandong Key Laboratory of Intelligent Electronic Packaging Testing and Application Shandong University Key Lab of Optical Astronomy National Astronomical ObservatoriesChinese Academy of Sciences University of Chinese Academy of Sciences School of Computer Science and Technology Taiyuan University of Science and Technology

The study of carbon-enhanced metal-poor (CEMP) stars is of great significance for understanding the chemical evolution of the early universe and stellar *** stars are characterized by carbon overabundance and are classified into several subclasses based on the abundance patterns of neutron-capture elements,including CEMP-s,CEMP-no,CEMP-r,and CEMP-r/*** subclasses provide important insights into the formation of thefirst stars,early stellar nucleosynthesis,and supernova ***,one of the major challenges in CEMP star research is the relatively small sample size of identified stars,which limits statistical analyses and hinders a comprehensive understanding of their ***,a series of large-scale spectroscopic survey projects have been launched and developed in recent years,providing unprecedented opportunities and technical challenges for the search and study of CEMP *** this end,this paper draws on the progress and future prospects of existing methods in constructing large CEMP data sets and offers an in-depth discussion from a technical standpoint,focusing on the strengths and *** addition,we review recent advancements in the identification of CEMP stars,emphasizing the growing role of machine learning in processing and analyzing the increasingly large data sets generated by modern astronomical *** to traditional spectral analysis methods,machine learning offers greater efficiency in handling complex data,automatic extraction of stellar parameters,and improved prediction *** these advancements,the research faces persistent challenges,including the scarcity of labeled samples,limitations imposed by low-resolution spectra,and the lack of interpretability in machine learning *** address these issues,the paper proposes potential solutions and future research directions aimed at advancing the study of CEMP stars and enhancing our understanding of their role in the chemical evolutio

关键词： methods: data analysis methods: observational stars: abundances stars: carbon surveys

来源：评论

学校读者我要写书评

暂无评论

Implementation of AI Collaboration in IOT-Enabled Systems for Better Response 4

Implementation of AI Collaboration in IOT-Enabled Systems fo...

引用

4th IEEE International Conference on Advance Computing and Innovative Technologies in engineering, ICACITE 2024

作者： Paul, M. Mano Raja Nagar, Mayank Rajan, T. Surulivel Badhoutiya, Arti Singh, K. Ranjith Choubey, Abhishek Nehru Institute of Engineering and Technology Department of Electrical and Electronic Engineering T.N India Ies College of Technology Department of Computer Science & Engineering M.P. Bhopal India Prince Shri Venkateshwara Padmavathy Engineering College Department of Mechanical Engineering TN India Gla University Department of Electrical Engineering Mathura India Karpagam Academy of Higher Education Department of Computer Science India Sreenidhi Institute of Science and Technology ghatkesar Hyderabad Yamnampet India

ISBN: (纸本)9798350360165

Artificial Intelligence (AI)-driven flexible methods play a crucial part in saving resources within the world of Internet about Things (IoT)-enabled smart healthcare devices. The tiny size and wireless data sharing capabilities of Body networked sensors (BSNs) have gained significant attention in current medical media technologies. The long-term feasibility of an effective health system hangs on the high vitality economy, communication dependability, and longer battery lives of personal sensing devices. Addressing the necessity to improve sensor-based wearable methods in BSNs across collecting information, healthcare tracking, and analysis, this work presents key advances. Firstly, a groundbreaking framework is suggested to simultaneously improve energy efficiency, storage lifetime, and dependability for smart and interactive healthcare. Secondly, the study proposes an Adaptive Transmission Info Rate (ATDR) method, improving energy usage by changing the active duration of the sense node over the continually changing wireless channel. Additionally, an Adaptable Routing System (SARA) is built, adding a dynamic origin routing system for an efficient in terms of energy and quickest path, breaking from conventional navigation methods. The trial setting uses real-time datasets, showing widespread and cost-effective treatment through smart devices. The results suggest that the proposed methods beat standards in terms of magnified energy savings, improved reliability, and extended battery lifetime for handheld devices. © 2024 IEEE.

关键词： Body sensor networks

来源：评论

学校读者我要写书评

暂无评论

Multifold Secured Bank Application Authentication Service Using Random Visual Cryptography and Multimodal Steganography with Blockchain technology 1

引用

3rd International Conference on Recent Trends in Machine Learning, IoT, Smart Cities and Applications, ICMISC 2022

作者： Jahnavi, S. Nandini, C. Computer Science and Engineering Department Dayananda Sagar Academy of Technology and Management Visvesvaraya Technological University Karnataka Bangalore560082 India

ISBN: (数字)9789811960888

ISBN: (纸本)9789811960871

DATA protection is becoming increasingly critical as the demand for online services grows, which biometrics authentication can help with. To communicate biometric data authenticating services over the Internet in a secure channel is essential. To ensure the security and the existence of the biometric information over the communication channel Random Visual Cryptography and Multimodal Mask Steganography is used. The biometric features captured are embedded using multimodal mask steganography, considering mask image and cover image. The resultant stego image shares are created using Random visual cryptography technique. The resultant shares are placed in blockchain using blockchain technology to improve security aspects of the cloud application and reliable transmission of data. Combining proposed Random Visual cryptography and multimodal mask steganography methods with blockchain technology ensures high reliable authentication system of server room access or banking application. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

Gesture-Based Virtual Keyboard: Enhancing Human-computer Interaction Through Hand Tracking

Gesture-Based Virtual Keyboard: Enhancing Human-Computer Int...

引用

2024 International Conference on Cybernation and Computation, CYBERCOM 2024

作者： Singh Yadav, Ajay Pratap Yadav, Ganesh Kumar Nagar, Sheetal Sharma, Brijesh Kumar Rawat, Daksh Bhakuni, Vihan Singh Uttar Pradesh Greater Noida201310 India JSS Academy of Technical Education Department of Computer Science and Engineering UP Noida201301 India IIMT College of Engineering Department of ECE Greater Noida India SRM Institute of Science and Technology Department of Computer Applications Modinagar India Graphic Era Hill University Computer Science and Engineering Dehradun India

ISBN: (纸本)9798331542344

In this study, an algorithm is developed that utilizes various OpenCV, cvzone and python libraries to provide a virtual keyboard. The program detects hand gestures, specifically the tips of fingers captured through a webcam, to simulate keystrokes corresponding to buttons on the virtual keyboard displayed on the screen, using Hand tracking module provided by Mediapipe. The virtual keyboard consists of keys of 'QWERTY' format and are represented by a graphical element. When the index finger approaches the corresponding key, it is highlighted and when distance between the tips of middle finger and index finger matches the required distance the corresponding keystroke is generated. The program captures video frames continuously from the webcam, identifies hand gestures, and accordingly updates the virtual keyboard's display. Additionally, the program outputs the text generated in a designated area of the screen using pynput library. The implementation uses object-oriented programming principles, with classes defined for buttons and hand detection functionality. © 2024 IEEE.

关键词： Object oriented programming

来源：评论

学校读者我要写书评

暂无评论

GLIMMER: Incorporating Graph and Lexical Features in Unsupervised Multi-Document Summarization 27

GLIMMER: Incorporating Graph and Lexical Features in Unsuper...

引用

27th European Conference on Artificial Intelligence, ECAI 2024

作者： Liu, Ran Liu, Ming Yu, Min Jiang, Jianguo Li, Gang Zhang, Dan Li, Jingyuan Meng, Xiang Huang, Weiqing Institute of Information Engineering Chinese Academy of Sciences China School of Cyber Security University of Chinese Academy of Sciences China School of Information Technology Deakin University Australia Department of Computer Science Beijing Technology and Business University China

ISBN: (纸本)9781643685489

Pre-trained language models are increasingly being used in multi-document summarization tasks. However, these models need large-scale corpora for pre-training and are domain-dependent. Other non-neural unsupervised summarization approaches mostly rely on key sentence extraction, which can lead to information loss. To address these challenges, we propose a lightweight yet effective unsupervised approach called GLIMMER: a Graph and LexIcal features based unsupervised Multi-docuMEnt summaRization approach. It first constructs a sentence graph from the source documents, then automatically identifies semantic clusters by mining low-level features from raw texts, thereby improving intra-cluster correlation and the fluency of generated sentences. Finally, it summarizes clusters into natural sentences. Experiments conducted on Multi-News, Multi-Xscience and DUC-2004 demonstrate that our approach outperforms existing unsupervised approaches. Furthermore, it surpasses state-of-the-art pre-trained multi-document summarization models (e.g. PEGASUS and PRIMERA) under zero-shot settings in terms of ROUGE scores. Additionally, human evaluations indicate that summaries generated by GLIMMER achieve high readability and informativeness scores. © 2024 The Authors.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Dual-loop joint torque control based on lower limb exoskeleton robots 18

Dual-loop joint torque control based on lower limb exoskelet...

引用

18th IEEE Conference on Industrial Electronics and Applications, ICIEA 2023

作者： Zhang, Meitong Zhang, Chi Yi, Feng Shu, Xindong Yuan, Ting Li, Zheng Hebei University of Science and Technology Ningbo Institute of Materials Technology and Engineering Chinese Academy of Sciences School of Electrical Engineering Ningbo315201 China Chinese Academy of Sciences Zhejiang Key Laboratory of Robotics and Intelligent Manufacturing Equipment Technology Ningbo Institute of Materials Technology and Engineering Ningbo315201 China University of Chinese Academy of Sciences Ningbo Institute of Materials Technology and Engineering Chinese Academy of Sciences College of Materials Science and Opto-Electronic Technology Ningbo315201 China Ningbo University Ningbo Institute of Materials Technology and Engineering Chinese Academy of Science Faculty of Electrical Engineering and Computer Science Ningbo315201 China Hebei University of Science and Technology School of Electrical Engineering Shijiazhuang050000 China

ISBN: (纸本)9798350312201

In the framework of impedance control, the torque trajectory of the joint is unpredictable due to nonlinear factors during the movement of the exoskeleton. To precisely predict command torque trajectory and compensate for the disturbances, we propose a dual-loop joint torque control system including predictive feedforward and a disturbance observer. A predictive feedforward module is adopted as the control outer loop based on the feedback control, and a general compensation module is used to compensate for the prediction error. The inner loop adopts a disturbance observer to estimate and compensate for low-frequency external disturbances. The simulation, which utilizes the actual torque of hip and knee joints as a reference, is conducted to prove the effectiveness of the proposed method. The comparison results demonstrate that the proposed control method can realize high bandwidth for exoskeleton joint control. © 2023 IEEE.

关键词： Torque control

来源：评论

学校读者我要写书评

暂无评论

Integrated Smart Patch for Heart Rate and Respiratory Rate Monitoring 14

Integrated Smart Patch for Heart Rate and Respiratory Rate M...

引用

14th International Conference on Measurement, MEASUREMENT 2023

作者： Gogola, Daniel Ackovska, Nevena Bagin, Richard Koteska, Bojana Bogdanova, Ana Madevska Kostoska, Magdalena Lehocki, Fedor Tysler, Milan Slovak Academy of Sciences Institute of Measurement Science Bratislava Slovakia Faculty of Computer Science and Engineering Skopje Macedonia Stu Faculty of Electrical Engineering and Information Technology Bratislava Slovakia

ISBN: (纸本)9788097262976

A wearable smart patch was designed to monitor the vital parameters of mass casualties' victims after the first triage. The device captures ECG, PPG, and respiration signals and triggers an alarm if the heart rate (HR) or respiration rate (RR) exceeds the specified limits and indicates a threat to the victim's life. To obtain a robust and reliable solution, the same parameters are derived from two or three independent signals. In this study, ECG signals have been recorded from different positions on the chest, and the performance of several algorithms for HR and RR extraction was tested. The initial measurements show that HR estimation is more accurate and reliable than RR estimation. The best results, considering both, the HR and RR calculations, were achieved when Pan-Tompkins's algorithm was used, and ECG electrodes were placed vertically on the right anterior chest. Increasing the length of the evaluated ECG signal above 30 seconds did not significantly affect the HR and RR calculation, regardless of the algorithm used. © 2023 Institute of Measurement science, Slovak academy o.

关键词： Electrocardiograms

来源：评论

学校读者我要写书评

暂无评论

A pretrained transformer model for decoding individual glucose dynamics from continuous glucose monitoring data

引用

National science Review 2025年第5期12卷 362-381页

作者： Yurun Lu Dan Liu Zhongming Liang Rui Liu Pei Chen Yitong Liu Jiachen Li Zhanying Feng Lei M.Li Bin Sheng Weiping Jia Luonan Chen Huating Li Yong Wang Center for Excellence in Mathematical Sciences National Center for Mathematics and Interdisciplinary Sciences Hua Loo-Keng Center for Mathematical Sciences Key Laboratory of Management Decision and Information System Academy of Mathematics and Systems ScienceChinese Academy of Sciences School of Mathematics University of Chinese Academy of SciencesChinese Academy of Sciences Department of Endocrinology and Metabolism Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of MedicineShanghai Diabetes InstituteShanghai Clinical Center for Diabetes Shanghai Key Laboratory of Diabetes Mellitus Key Laboratory of Systems Health Science of Zhejiang Province School of Life ScienceHangzhou Institute for Advanced StudyUniversity of Chinese Academy of Sciences BGI-Research School of Mathematics South China University of Technology Department of Statistics Department of Biomedical Data Science Bio-X ProgramStanford University Department of Computer Science and Engineering Shanghai Jiao Tong University State Key Laboratory of Cell Biology Center for Excellence in Molecular Cell Science Shanghai Institute of Biochemistry and Cell Biology Chinese Academy of Sciences Guangdong Institute of Intelligence Science and Technology Pazhou Laboratory (Huangpu)

Continuous glucose monitoring(CGM) technology has grown rapidly to track real-time blood glucose levels and trends with improved sensor accuracy. The ease of use and wide availability of CGM will facilitate safe and effective decision making for diabetes management. Here, we developed an attention-based deep learning model, CGMformer, pretrained on a well-controlled and diverse corpus of CGM data to represent individual's intrinsic metabolic state and enable clinical applications. During pretraining, CGMformer encodes glucose dynamics including glucose level, fluctuation, hyperglycemia, and hypoglycemia into latent space with self-supervised learning. It shows generalizability in imputing glucose value across five external datasets with different populations and metabolic states(MAE = 3.7 mg/d L). We then fine-tuned CGMformer towards a diverse panel of downstream tasks in the screening of diabetes and its complications using task-specific data, which demonstrated a consistently boosted predictive accuracy over direct fine-tuning on a single task(AUROC = 0.914 for type 2 diabetes(T2D) screening and 0.741 for complication screening). By learning an intrinsic representation of an individual's glucose dynamics,CGMformer classifies non-diabetic individuals into six clusters with elevated T2D risks, and identifies a specific cluster with lean body-shape but high risk of glucose metabolism disorders, which is overlooked by traditional glucose measurements. Furthermore, CGMformer achieves high accuracy in predicting an individual's postprandial glucose response with dietary modelling(Pearson correlation coefficient = 0.763)and helps personalized dietary recommendations. Overall, CGMformer pretrains a transformer neural network architecture to learn an intrinsic representation by borrowing information from a large amount of daily glucose profiles, and demonstrates predictive capabilities fine-tuned towards a broad range of downstream applications, holding promise for the ear

关键词： pretrained model diabetes continuous glucose monitoring glucose dynamics

来源：评论

学校读者我要写书评

暂无评论

Artificial Intelligence-Driven Power Management System for Enhanced Efficiency in Smart Grids

Artificial Intelligence-Driven Power Management System for E...

引用

2024 IEEE International Conference on Recent Advances in science and engineering technology, ICRASET 2024

作者： Deepak, B.L. Sabin, T.T. Darshana, A. Vindya, L. Suma Manjunath, N. Mallikarjunaswamy, S. Shilpa, M. Rekha, V. Sjc Institute of Technology Dept of Information Science and Engineering Chickballapur India Sjc Institute of Technology Dept of Ai & Ds Chickballapur India Jss Academy of Technical Eduction Dept of Electronics and Communication Engineering Bengaluru India Bengaluru Institute of Technology Dept of Information Science and Engineering Bengaluru India Christ University Department of Computer Science and Engineering Bangalore India

ISBN: (纸本)9798350388602

Smart grids are crucial in modern energy systems, integrating sensors, communication networks, and control mechanisms to improve energy distribution. However, conventional methods like Time-Based Scheduling (TBS), Demand Response Mechanism (DRM), and Static Load Distribution (SLD) struggle with inefficiencies such as poor real-time load balancing, limited adaptability, and inaccurate power usage predictions. These drawbacks hinder the overall effectiveness of smart grid *** address these challenges, the proposed AI-Driven Power Management System (AI-PMS) leverages machine learning to optimize energy distribution and predict demand more accurately. By using historical and real-time data, AI-PMS dynamically adjusts power flow, ensuring better load distribution and efficiency. This method surpasses the limitations of TBS, DRM, and SLD. The implementation of AI-PMS has resulted in a 0.25% improvement in energy efficiency and a 0.15% boost in grid reliability, significantly reducing power losses. The system also enhances the integration of renewable energy sources, making it a more sustainable and adaptive solution for modern smart grids. These advancements promise better power management and lower operational costs for grid operators. © 2024 IEEE.

关键词： Smart power grids

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：