A5072915694- Cloud Computing and Resource Management
- Distributed and Parallel Computing Systems
- IoT and Edge/Fog Computing
- Data Stream Mining Techniques
- Cloud Data Security Solutions
- Distributed systems and fault tolerance
- Graph Theory and Algorithms
- Privacy-Preserving Technologies in Data
- Advanced Data Storage Technologies
- Advanced Database Systems and Queries
- Opportunistic and Delay-Tolerant Networks
- Data Management and Algorithms
- Parallel Computing and Optimization Techniques
- Advanced Neural Network Applications
- Advanced Image and Video Retrieval Techniques
- Security and Verification in Computing
- Peer-to-Peer Network Technologies
- Cryptography and Data Security
- Privacy, Security, and Data Protection
- Mobile Ad Hoc Networks
- Robotics and Sensor-Based Localization
- Domain Adaptation and Few-Shot Learning
- Data Mining Algorithms and Applications
- Machine Learning and ELM
- Access Control and Trust
University of Illinois Urbana-Champaign
2022-2024
China University of Geosciences (Beijing)
2015-2024
ZheJiang Academy of Agricultural Sciences
2024
Ministry of Agriculture and Rural Affairs
2024
Beijing Institute of Big Data Research
2020
Intel (United States)
2020
Tsinghua University
2013-2019
Ministry of Natural Resources
2019
State Key Laboratory on Integrated Optoelectronics
2019
Chinese Academy of Sciences
2019
Establishing correspondences between two images requires both local and global spatial context. Given putative of feature points in views, this paper, we propose Order-Aware Network, which infers the probabilities being inliers regresses relative pose encoded by essential matrix. Specifically, proposed network is built hierarchically comprises three novel operations. First, to capture context sparse correspondences, clusters unordered input learning a soft assignment These are canonical...
Cloud computing is still in its infancy spite of gaining tremendous momentum recently, high security one the major obstacles for opening up new era long dreamed vision as a utility. As sensitive applications and data are moved into cloud centers, run on virtual resources form machine. This unique attributes, however, poses many novel tangible intangible challenges. It might be difficult to track issue environments. So this paper primarily aims highlight security, privacy trust issues current...
Convolutional Neural Networks (CNNs) have become deeper and more complicated compared with the pioneering AlexNet. However, current prevailing training scheme follows previous way of adding supervision to last layer network only propagating error information up layer-by-layer. In this paper, we propose Deeply-supervised Knowledge Synergy (DKS), a new method aiming train CNNs improved generalization ability for image classification tasks without introducing extra computational cost during...
Today, one of the most important factors for success cloud computing is to create trust and security. Cloud will face a lot challenges when key element absent. There are no special evaluation models environment. In this paper, definition in systems introduced properties analyzed. Based on semantics trust, an extensible model named ETEC proposed, which includes time-variant comprehensive method expressing direct space-variant calculating recommendation trust. To compute systems, algorithm...
Establishing correspondences between two images requires both local and global spatial context. Given putative of feature points in views, this paper, we propose Order-Aware Network, which infers the probabilities being inliers regresses relative pose encoded by essential matrix. Specifically, proposed network is built hierarchically comprises three novel operations. First, to capture context sparse correspondences, clusters unordered input learning a soft assignment These are canonical...
Security of cloud computing is one the challenges to be addressed before novel pas-as-you-go business model widely applied, and dependability most important means improve security current heterogeneous platforms. Previous research on in systems only uses qualitative approaches there are few systematic works systems. In this paper, definition given a series quantitative indicators presented evaluate dependability. A CDSV established enhance environments. System-level virtualization techniques...
To allocate cloud resources efficiently and obtain the maximum economic benefit are major goals of resource providers users. The objective this paper is to present a novel allocation algorithm named NECDA overcome some shortcomings current mechanisms. A model m*n type based on M/M/1 queuing system established first. Then Nash equilibrium mode applied in computing among environment, optimization each performance-QoS economic-QoS, achieved by taking advantage continuous double auction step,...
The issue of high system stability is one the major obstacles for real-time computing over fluctuating big data streams. A stable scheduling more important than an efficient stream applications, especially when a to be rescheduled dynamically at runtime. In this paper, online strategy with makespan guarantee SOMG discussed, which includes following features: 1) profiling mathematical relationships between stability, response time, and resource utilization, indicating conditions meet...
High security of cloud computing is one the most challenges to be addressed before novel pas-as-you-go business paradigm widely applied over internet. Trust brings a means improve platforms. However, there are few systematic works on trust in systems. In this paper, definition systems given and properties analysed by referring fruits from social science. Based semantics trust, dynamic multi-dimensional model named DMTC proposed, which includes time-variant comprehensive evaluation method for...
In the research on mobile sensor networks, coverage control is one of most important challenges. For network constructed by random distribution, better can be achieved topology adjustment utilizing mobility nodes. This paper investigates how to make use radius and nodes improve coverage. A node distribution optimization scheme based multi-objective differential evolution algorithm proposed. Simulation results show that proposed quickly achieve a network, increase effective rate, reduce...
High fault tolerance issue is one of the major obstacles for opening up new era high serviceability cloud computing as plays a key role in order to ensure serviceability. In most current clouds, check pointing, process saving application states, and replication, replicating hot data, usually stable storage, have been two common strategies. However, when, where, how often insert pointing or replicate data become challenges are ignored clouds. this paper, definitions fault, error, failure...
Entity–relationship extraction plays a pivotal role in the construction of domain knowledge graphs. For aquatic disease domain, however, this relationship is formidable task because overlapping relationships, data specialization, limited feature fusion, and imbalanced samples, which significantly weaken extraction’s performance. To tackle these challenges, study leverages published books websites as sources to compile text corpus, establish datasets, then propose Fd-CasBGRel model...