检索结果-内蒙古大学图书馆

YOLO-SIFD:YOLO with Sliced Inference and Fractal Dimension Analysis for Improved Fire and Smoke Detection

Computers, Materials & Continua 2025年第3期82卷 5343-5361页

作者： Mariam Ishtiaq Jong-Un Won Artificial Intelligence Railroad Research Department Korea Railroad Research Institute(KRRI)Uiwang16105Republic of Korea Transportation System Engineering University of Science and Technology(UST)Daejeon34113Republic of Korea

Fire detection has held stringent importance in computer vision for over half a *** development of early fire detection strategies is pivotal to the realization of safe and smart cities,inhabitable in the ***,the development of optimal fire and smoke detection models is hindered by limitations like publicly available datasets,lack of diversity,and class *** this work,we explore the possible ways forward to overcome these challenges posed by available *** study the impact of a class-balanced dataset to improve the fire detection capability of state-of-the-art(SOTA)vision-based models and propose the use of generative models for data augmentation,as a future work ***,a comparative analysis of two prominent object detection architectures,You Only Look Once version 7(YOLOv7)and YOLOv8 has been carried out using a balanced dataset,where both models have been evaluated across various evaluation metrics including precision,recall,and mean Average Precision(mAP).The results are compared to other recent fire detection models,highlighting the superior performance and efficiency of the proposed YOLOv8 architecture as trained on our balanced ***,a fractal dimension analysis gives a deeper insight into the repetition of patterns in fire,and the effectiveness of the results has been demonstrated by a windowing-based inference *** proposed Slicing-Aided Hyper Inference(SAHI)improves the fire and smoke detection capability of YOLOv8 for real-life applications with a significantly improved mAP performance over a strict confidence ***8 with SAHI inference gives a mAP:50-95 improvement of more than 25%compared to the base YOLOv8 *** study also provides insights into future work direction by exploring the potential of generative models like deep convolutional generative adversarial network(DCGAN)and diffusion models like stable diffusion,for data augmentation.

关键词： Fire detection smoke detection class-balanced dataset you only look once(YOLO) slicing-aided hyper inference(SAHI) fractal dimension generative adversarial network(GAN) diffusion models

来源：评论

学校读者我要写书评

暂无评论

Recognition of Typical Highway Driving Scenarios for Intelligent Connected Vehicles Based on Long Short-Term Memory Network

Recognition of Typical Highway Driving Scenarios for Intelli...

引用

11th International Conference on Vehicle Technology and Intelligent Transport systems, VEHITS 2025

作者： Feng, Xinjie Yang, Shichun Peng, Zhaoxia Chen, Yuyi Sun, Bin Lu, Jiayi Wang, Rui Cao, Yaoguang School of Transportation Science and Engineering Beihang University Beijing China State Key Lab of Intelligent Transportation System Beihang University Beijing China Hangzhou International Innovation Institute Beihang University Hangzhou China

ISBN: (纸本)9789897587450

In the complex traffic environment where intelligent connected vehicles (ICVs) and traditional vehicles coexist, accurately identifying the driving scenarios of a vehicle helps ICVs make safer and more efficient decisions, while also enabling performance evaluation across different scenarios to further optimize system capabilities. This paper presents a typical highway driving scenarios recognition model with extensive scenario coverage and high generalizability. The model first categorizes the constituent elements of driving scenarios and extracts the core elements of typical highway scenarios. Then, based on a long short-term memory (LSTM) network architecture, it extracts features from the ego vehicle and surrounding vehicles to identify the typical driving scenarios in which the ego vehicle is located. The model was tested and validated on the HighD dataset, achieving an overall accuracy of 96.74% for four typical highway scenarios: Lane-change, Car-following, Alongside vehicle cut-in, and Preceding vehicle cut-out. Compared to baseline models, the proposed model demonstrated superior performance. Copyright © 2025 by SCITEPRESS - science and Technology Publications, Lda.

关键词： Advanced traffic management systems

来源：评论

学校读者我要写书评

暂无评论

Prospects on mangrove knowledge services in the smart era:From plant atlas to knowledge graphs

引用

science China Earth sciences 2025年第1期68卷 111-127页

作者： Zhi-Wei HOU Wenlong JING Cheng-Zhi QIN Ji YANG Qing XIA Xiaoling YIN Key Laboratory of Guangdong for Utilization of Remote Sensing and Geographical Information System Guangdong Open Laboratory of Geospatial Information Technology and ApplicationGuangzhou Institute of GeographyGuangdong Academy of SciencesGuangzhou510070China Southern Marine Science and Engineering Guangdong Laboratory(Guangzhou) Guangzhou511458China State Key Laboratory of Resources and Environmental Information System Institute of Geographic Sciences and Natural Resources ResearchChinese Academy of SciencesBeijing100101China College of Resources and Environment University of Chinese Academy of SciencesBeijing100049China School of Traffic&Transportation Engineering Changsha University of Science and TechnologyChangsha410114China

Mangroves are crucial to the ecological security of the Earth and human *** management,conservation,and restoration are of great importance and necessitate the support of spatio-temporal information and multidisciplinary knowledge such as biology and *** knowledge services such as plant atlas provide illustrated textual knowledge of ***,this kind of service is oriented to information retrieval and is incapable of effectively mining and utilizing fragmented knowledge from multi-source heterogeneous data,facing the problem of“massive data,rare knowledge”.Knowledge graphs are capable of extracting,organizing,and fusing the knowledge contained in massive data into semantic networks that can be understood and computed by *** provide a solution for the realization of intelligent knowledge *** on the urgent need for mangrove knowledge acquisition,formal representation,and intelligent services,this paper proposes a research prospect on mangrove knowledge graphs and knowledge *** first analyze the similarities and differences between various domain-specific concepts of *** this basis,we define the mangrove knowledge graph as a large-scale knowledge base that integrates multi-disciplinary knowledge and spatio-temporal information with mangrove ecosystems as the ***,we propose a research framework for mangrove knowledge services that can realize the transformation from multi-modal data to intelligent knowledge services,including multiple research levels such as ubiquitous data sensing and aggregation,knowledge organization and graph construction,and intelligent mangrove knowledge ***,the methods and workflow for constructing mangrove knowledge graphs are ***,we discuss the challenges and possible future directions of mangrove knowledge services in the smart era,including the construction of a mangrove knowledge system that integrates the domain-specific charac

关键词： Mangrove Tupu Knowledge graph Multi-modal Knowledge service

来源：评论

学校读者我要写书评

暂无评论

Wideband Channel Measurement and Characterization Based on 5G Passive Sounding in Railway Scenarios

引用

China Communications 2025年第4期22卷 13-26页

作者： Zheng Peng Fei Dan Chen Chen Chen Haoran Huang Yanyan Ai Bo School of Electronic and Information Engineering Beijing Jiaotong UniversityBeijing 100044China State Key Laboratory of Advanced Rail Autonomous Operation Beijing Jiaotong UniversityBeijing 100044China Nanjing Rongcai Transportation Technology Research Institute Co. LtdNanjing 211121China Frontiers Science Center for Smart High-Speed Railway System Beijing Jiaotong UniversityBeijing 100044China

With the development of wireless communication,the fifth generation mobile communication technology(5G)has emerged as a hot topic in highspeed railway communication system and has moved towards industrial *** the radio propagation characteristics in 5G high-speed train(HST)scenarios is essential for enhancing wireless coverage and overall system *** propose a novel 5G passive sounding scheme to extract channel impulse responses(CIRs)using channel state information reference signals(CSI-RS)from the target 5G base station(BS).Detailed procedures for timefrequency synchronization,CSI-RS detection and extraction are presented through *** the laboratory work involving absolute power calibration,phase coherence calibration and power delay profile(PDP)validation,we validate the accuracy and performance of the developed ***,a measurement campaign was conducted in HST scenarios encompassing both residential and undeveloped *** path loss(PL)model and the channel characteristics including stationarity interval(SI),multipath components(MPCs),shadow fading(SF),Rician K-factor,root mean square(RMS)delay spread and received correlation coefficients are analyzed and *** estimated channel characteristics and the statistical model presented in this paper will contribute to the research on HST radio propagation and the development of 5G railway communication systems.

关键词： 5G channel characterization channel measurement high-speed train passive sounding

来源：评论

学校读者我要写书评

暂无评论

Speed Planning for Integrated Eco-Driving and Bus Bunching Mitigation in Connected Electric Buses

引用

IEEE Transactions on transportation Electrification 2025年

作者： Han, Yu Ma, Xiaolei Li, Xin Bian, Jing Wang, Wenwei Jiang, Shengchuan Beihang University School of Transportation Science and Engineering Key Laboratory of Intelligent Transportation Technology and System of the Ministry of Education Beijing102206 China Dalian Maritime University College of Transportation Engineering Dalian116026 China Beijing Institute of Technology College of Mechanical Engineering Beijing102206 China University of Shanghai for Science and Technology College of Traffic Engineering Business School Shanghai200090 China

Eco-driving and bus bunching are two major challenges for connected electric buses (CEBs). Eco-driving aims to minimize energy consumption, whereas bus bunching occurs when consecutive buses arrive at the same station simultaneously. Existing research rarely addresses both issues together. To address this gap, this paper proposes a novel speed planning approach that addresses both problems via nonlinear model predictive control (NMPC) and imitation learning. A multi-objective NMPC model is developed that considers practical factors such as energy consumption, time headway deviation, and traffic lights. To save computational resources, a speed planning network (SPN) based on transformer and long short-term memory (LSTM) architectures is designed to mimic the NMPC planner. Additionally, a knowledge distillation method is introduced to reduce the SPN's memory footprint by incorporating mixed knowledge. Extensive experiments show that the NMPC model ensures nonstop passage through intersections and performs better across multiple metrics, including energy consumption and time headway deviation, than several baselines do. The SPN achieves similar performance to NMPC while significantly improving real-time efficiency, and the proposed distillation method further reduces memory usage while maintaining acceptable performance. © 2015 IEEE.

关键词： Predictive control systems

来源：评论

学校读者我要写书评

暂无评论

Self-Triggered Stochastic MPC With Adaptive Prediction Horizon for Cloud-Based Connected Vehicles Subject to Chance Constraints

引用

IEEE Transactions on Vehicular Technology 2025年

作者： Chen, Jicheng Zhang, Hui Beihang University School of Reliability and System Engineering Beijing100191 China Beihang University Hangzhou International Innovation Institute Hangzhou311115 China Beihang University School of Transportation Science and Engineering Beijing100191 China

This work presents a novel self-triggered stochastic model predictive control (SMPC) scheme for cloud-based vehicle path following control system subject to chance constraints. First, we model the cloud-based vehicle path following control system from a networked stochastic control system perspective. Unlike the conventional periodical sampling approach, a self-triggered mechanism (STM) with adaptive prediction horizon is developed to determine the next sampling time instant and inter-sampling control inputs at each sampling time instant. This mechanism can efficiently reduce the data transmission frequency in the vehicle-cloud communication network, leading to a lower communication load and thus improving the reliability of the system. The STM comprises a set of optimization problems with an adaptive prediction horizon. The optimization problems and threshold design explicitly take the vehicle-cloud communication load into account. Furthermore, a stochastic model predictive control problem with modified constraint tightening and terminal constraint is defined by considering the influence of STM. We develop sufficient conditions to guarantee the closed-loop chance constraints satisfaction in the presence of both adaptive STM and additive disturbances. Then, the recursive feasibility of the optimal control problem and closed-loop stability of the system are investigated. Finally, we illustrate the benefits and effectiveness of the proposed method through numerical examples in vehicle path following control problem. © 1967-2012 IEEE.

关键词： Stochastic control systems

来源：评论

学校读者我要写书评

暂无评论

Data-Driven Handling Qualities Assessment for Urban Air Mobility Vehicles Using Physiological Signals

SSRN

引用

SSRN 2025年

作者： Zhang, Shuguang Wu, Yibing Zintl, Michael Holzapfel, Florian Huang, Jun Li, Yuhan School of Aeronautic Science and Engineering Beihang University Beijing100191 China Institute of Flight System Dynamics Technical University of Munich Munich80333 Germany School of Transportation Science and Engineering Beihang University Beijing100191 China

The rapid development of Urban Air Mobility (UAM) vehicles necessitates a reliable and efficient method for assessing handling qualities (HQ) in a wide range of operators. The Cooper-Harper Rating (CHR) method relies heavily on subjective evaluations by highly trained test pilots, which can be time-consuming, costly, and prone to individual biases. This study addresses the limitations of CHR by proposing a novel approach that utilizes quanti-tative metrics of pilot compensation and deep learning techniques to estimate HQ. A flight test was conducted to collect and extract physiological metrics from both expert test pilots and average pilots, with expert test pilots’ data guiding the selection of the most responsive indicators based on copula entropy to serve as quantitative metrics. These quantitative metrics from expert test pilots were then used to train a convolutional neural network (CNN)-based model. The model demonstrated strong performance, achieving an accuracy of 96.15% and showing robust generalization to average pilots with an accuracy of 85.21% through fine-tuning. This approach not only complements the CHR but also reduces reliance on highly trained test pilots, providing a scalable and efficient solution for broader UAM applications. This study contributes to the field by assessing HQ in a data-driven manner and offering a promising tool for UAM vehicle testing, reducing reliance on subjective evaluations and addressing the challenges of scaling HQ assessments to diverse pilot groups. © 2025, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Research on New Type Secondary Inerter Suspension Damping Devices for Rail Vehicles Based on Low Frequency and High Frequency Track Excitation 3

Research on New Type Secondary Inerter Suspension Damping De...

引用

3rd International Conference on Rail transportation, ICRT 2024

作者： Dilai, C. Gengchen, S. Chenxu, L. Rui, S. Lixia, S. School of Railway Transportation Shanghai Institute of Technology Shanghai China China Railway Test and Certification Center Limited Beijing China Standards and Metrology Research Institute of China Academy of Railway Sciences Corporation Limited Beijing China Railway Science and Technology Research and Development Center of China Academy of Railway Sciences Corporation Limited Beijing China Wheel-Rail System Laboratory National Engineering Research Center of System Technology for High-Speed Railway and Urban Rail Transit Beijing China

ISBN: (数字)9780784485941

ISBN: (纸本)9780784485941

Different types of secondary inerter suspension vibration damping devices are established set on the Matlab platform and compare and analyzes the dynamic indicators of vehicle operation, selecting the secondary inertial suspension damping device that significantly optimizes both the quality of vehicle operation and the suppression of low-frequency and high-frequency body sway. The research results indicate that for vehicle low-frequency swaying, all four types of new dual-system inertial damping suspension devices have an optimizing effect, with Condition S5 showing the most significant improvement. Compared to the traditional suspension damping device's lateral acceleration of 1.62 m/s2, it has been optimized by 57%, achieving 0.7 m/s2. For vehicle high-frequency swaying, Condition S5 again showing the most notable improvement. Relative to the traditional suspension damping device's lateral acceleration of 0.87 m/s2, it has been optimized by 52%, reaching 0.41 m/s2. After optimizing the parameters for various conditions, the results indicate that Condition S5, with an inertial coefficient of 5,000 kg, provides the best optimization effect on the stability of the vehicle's operation. © ASCE.

关键词： Automobile suspensions

来源：评论

学校读者我要写书评

暂无评论

HeadSonic: Usable Bone Conduction Earphone Authentication via Head-conducted Sounds

引用

IEEE Transactions on Mobile Computing 2025年

作者： He, Zhixiang Chen, Jing He, Kun Gu, Yangyang Deng, Qiyi Zhang, Zijian Du, Ruiying Zhao, Qingchuan Wu, Cong Wuhan University School of Cyber Science and Engineering Wuhan430072 China State Key Lab of Intelligent Transportation System Beijing China Beijing Institute of Technology School of Cyberspace Science and Technology China City University of Hong Kong Department of Computer Science Hong Kong The University of Hong Kong Department of Electrical and Electronic Engineering Hong Kong

Earables (ear wearables) are rapidly emerging as a new platform encompassing a diverse of personal applications, prompting the development of authentication schemes to protect user privacy. Existing earable authentication methods are all specifically designed for air-conduction earphones, which are not suited for bone conduction earphones (BCEs) that rely on bone conduction mechanisms. In this paper, we propose HeadSonic, a usable BCE authentication system based on the unique head-conducted sounds, which can be acquired when the user wears the BCE device. Specifically, the system emits a millisecond-level sound to initiate the authentication session. The signal captured by the BCE microphone is propagated through the user’s head, which is unique in density, geometry, and bone-tissue ratio. It operates implicitly, while maintaining robustness across different behaviors. Extensive experiments involving 60 subjects demonstrate that HeadSonic achieves a commendable balanced accuracy of 96.59%, proving its efficacy and resilience against replay and synthesis attacks. Our dataset and source codes are available at https://***/r/HeadSonic1CE4. © 2025 IEEE. All rights reserved.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

CRPark: A Smart Parking system with Competition and Reservation Services

引用

IEEE Transactions on Vehicular Technology 2025年

作者： Hu, Ziyi Cao, Yue Huang, Jie Wu, Libing Xu, Lexi Zhang, Ronghui Wuhan University School of Cyber Science and Engineering Wuhan430072 China Suzhou Research Institute of Wuhan University Suzhou215123 China China Unicom Research Institute Beijing100048 China Sun Yat-sen University Guangdong Key Laboratory of Intelligent Transportation System School of Intelligent Systems Engineering Guangdong 510275 China

With the help of communication technology and intelligent transportation system, smart parking has been further upgrading. Compared with traditional parking, smart parking alleviates users from finding available parking slots, through notifying users of available slot locations in advance. However, existing schemes on non-reservation smart parking may cause parking competition, because users will compete for the common known available slots. Although the reservation-based smart parking alleviates parking competition through reserving at slot for users in advance, it may cause parking difficulty when there are no available slots for reserving. A comprehensive parking service is urgently needed, to reduce parking competition under non-reservation smart parking and parking difficulty under reservation-based smart parking. Hence, we propose the CRPark, to manage both the reservation and competition smart parking services. Through the Gaussian Mixture model and heuristic algorithm, we determine the optimal parking lot reservation management strategy, to reduce both parking competition and parking difficulty. Besides, the parking competition analysis helps to estimate the parking cost accurately, so that parking cost can be reduced. Through simulation experiments, our scheme outperforms other schemes on reducing parking cost and balancing parking demand. © 1967-2012 IEEE.

关键词： Parking lot management Parking reservation Smart parking

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：