A5100330989- Automated Road and Building Extraction
- Energy Efficient Wireless Sensor Networks
- Underwater Vehicles and Communication Systems
- Human Mobility and Location-Based Analysis
- Geographic Information Systems Studies
- Indoor and Outdoor Localization Technologies
- Video Surveillance and Tracking Methods
- Data Management and Algorithms
- Traffic Prediction and Management Techniques
- Diverse Aspects of Tourism Research
- Remote Sensing and LiDAR Applications
- Transportation and Mobility Innovations
- Coal Properties and Utilization
- Asian American and Pacific Histories
- Distributed Control Multi-Agent Systems
- Migration, Aging, and Tourism Studies
- Advanced Wireless Communication Technologies
- Icing and De-icing Technologies
- Coal and Its By-products
- Asian Culture and Media Studies
- Energy Harvesting in Wireless Networks
- Access Control and Trust
- Geoscience and Mining Technology
- Gait Recognition and Analysis
- Security in Wireless Sensor Networks
Hefei University of Technology
2022-2025
Tongji University
2018-2022
Shanghai Artificial Intelligence Laboratory
2022
Beijing University of Posts and Telecommunications
2022
Chang'an University
2022
Meta (United States)
2013-2015
Menlo School
2013-2015
Chongqing University
2014
Lawrence Technological University
2014
Hewlett-Packard (United States)
2013
The inaccuracy of manually created digital road maps is a persistent problem, despite their high economic value. We present CrowdAtlas, which automates map update based on people's travels, either individually or crowdsourced. Its mobile navigation app detects significant portions GPS traces that do not conform to the existing map, as determined by state-of-the-art Viterbi matching. When there sufficient evidence collected, inference algorithms can automatically map. CrowdAtlas server...
With the recent advances of Convolutional Neural Networks (CNN) in computer vision, there have been rapid progresses extracting roads and other features from satellite imagery for mapping purposes. In this paper, we propose a new method road extraction using stacked UNets with multiple output. A hybrid loss function is used to address problem unbalanced classes training data. Post-processing methods, including map vectorization shortest path search hierarchical thresholds, help improve...
Digital road maps have become essential to many aspects of our lives. Unfortunately, they persistent quality issues, both in developing countries as well developed countries, evidenced by the recent Apple-Google map war. A survey British drivers showed 26% been directed their GPS go into no-entry areas, and news periodically reports car accidents caused or related digital maps. In addition correcting existing errors, need be frequently updated reflect latest constructions, closures,...
Journal Article Mining Large-Scale GPS Streams for Connectivity Refinement of Road Maps Get access Yanmin Zhu, Zhu * 1Shanghai Key Lab Scalable Computing and Systems, Shanghai, China2Department Computer Science Engineering, Shanghai Jiao Tong University, China *Corresponding author: yzhu@cs.sjtu.edu.cn Search other works by this author on: Oxford Academic Google Scholar Yin Wang, Wang 3Facebook, Menlo Park, USA George Forman, Forman 4HP Labs, Palo Alto, Hong Wei The Journal, Volume 58, Issue...
Because the galloping of iced conductors is one main disasters in State Grid, resulting huge economic and property losses every year, research on relevant monitoring methods great significance. The existing technology mainly based contact detection method, which presents potential electrical hazards power supply problems. In this paper, a conductor method target infrared sources put forward to overcome shortcomings methods. other words, an source label installed spacer, high-definition night...
In underwater sensor networks, nodes are usually sparsely arranged, and autonomous vehicles (AUVs) often used to collect data by moving. However, it is a key problem for AUVs shorten the path as much possible guarantee whole-network acquisition. Aiming solve this problem, paper proposes an AUV collection method based on affinity propagation clustering. First, clustering theory select with large residual energy cluster head nodes, results dynamically adjusted according communication radius of...
As people increasingly rely on road maps in the digital age, manually maintained cannot keep up with demand for accuracy and freshness, evidenced by recent iOS map incident bidding war Waze. There are many research works automatic inference using GPS data, some have suggested that Google Waze automate their update processes to degree user data. However, existing published work focuses refining geometry. In reality, connectivity issues at intersections, including missing connections unmarked...
以热重实验中的煤氧化增重现象为线索, 对400 ℃前煤氧化的反应机理和动力学过程进行了理论研究. 借助于平行反应理论, 将煤低中温氧化过程简化为3 个平行的竞争反应, 包括水分析出、氧吸附和固相氧化产物分解的反应以及直接的"burn-off"反应. 通过对不同氧体积浓度(10%, 21%和40%)下的实验数据全过程上的非线性最小二乘法拟合, 确定出各反应的动力学参数以及相应的温度作用范围. 模化的结果与实验数据高度贴近, 证明了反应模型和计算方法的可靠性. 煤低中温氧化过程中的质量变化, 主要是由氧吸附反应和固相氧化产物的分解反应所控制的; 由直接的"burn-off"反应所引起的质量变化则相对较小. 煤氧化增重现象对应的反应及动力学过程并不能由单一反应模型来近似. 氧浓度的提高导致氧吸附反应和固相氧化产物的分解反应加剧, 而"burn-off"反应受氧浓度的影响并不显著, 这是由用于解析动力学参数而选定的实验数据的温度范围和特征所决定的.
This paper puts forward an alternative method which deals with the simulation and estimation of urban traffic network. The is based on Cell Transmission Model (CTM), a macroscopic model for freeway modeling. Some modifications basic CTM are made to meet requirement In order test performance this model, small virtual network set up its results compared those Paramics', microscopic simulator implemented under similar conditions. Furthermore, much larger real metropolitan system constructed...
The ground traffic demand of multi-beam satellites is growing rapidly, but resources such as satellite power and bandwidth are limited. Beam hopping (BH) alleviates the problem insuffident transmitters, how to improve utilization limited increase throughput still a challenge in future. In this paper, starting from beam hopping, allocation systems analyzed first. results show that effect full frequency reuse works better than partial system with fewer beams, former can reduce complexity...
To address the problems of node deployment schemes in existing underwater sensor networks that lack consideration network connectivity and high costs, this article constructs an optimization model maximizes coverage minimizes costs while ensuring full connectivity. For NP-hard property model, improved moth flame algorithm based on fuzzy operators (IMF 2 O ) is proposed. First, comprehensively considering two performance metrics connectivity, a multi-objective selection mechanism proposed to...
Underwater pollution incidents occur frequently, and obtaining accurate information about their exact location real-time situation is helpful for promptly formulating plans to contain mitigate the situation. Autonomously adjusting position of sensors optimal coverage monitoring regions interest (e.g., oil spill zones, chemical contamination areas) in real time a significant challenge. To this end, article proposes hunting-style underwater sensor deployment based on level set method. This...