Wei Lv

ORCID: 0000-0001-8859-8972
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Cloud Computing and Resource Management
  • Machine Learning and Algorithms
  • Atmospheric chemistry and aerosols
  • Air Quality and Health Impacts
  • Visual and Cognitive Learning Processes
  • Mind wandering and attention
  • Digital Imaging for Blood Diseases
  • Face and Expression Recognition
  • Recycling and utilization of industrial and municipal waste in materials production
  • Advanced Graph Neural Networks
  • Atmospheric aerosols and clouds
  • Gait Recognition and Analysis
  • Distributed and Parallel Computing Systems
  • Cell Image Analysis Techniques
  • Human Pose and Action Recognition
  • Metallurgy and Material Forming
  • Combustion and Detonation Processes
  • Image Enhancement Techniques
  • Industrial Technology and Control Systems
  • Bioinformatics and Genomic Networks
  • Advanced Multi-Objective Optimization Algorithms
  • Graph Theory and Algorithms
  • Microstructure and mechanical properties
  • Tailings Management and Properties
  • Advanced Vision and Imaging

Northeastern University
2019-2024

Huazhong University of Science and Technology
2011-2017

Harbin University of Science and Technology
2017

Wuhan National Laboratory for Optoelectronics
2017

Peking University
2016

State Key Joint Laboratory of Environment Simulation and Pollution Control
2016

Shanghai Institute of Microsystem and Information Technology
2015

Beijing Normal University
2015

As an initial channel for users learning about a mobile game, the interactive experience of navigation interface will directly affect first impression on game and their subsequent behaviors willingness to use. This study aims investigate players’ visual attention mechanisms various levels games’ interfaces under free-browsing task-oriented conditions. Eye-tracking glasses questionnaire were used measure games. The results show that in condition, fixation count, saccade count average...

10.3390/app9163242 article EN cc-by Applied Sciences 2019-08-08

In order to solve the problems of small capacity structured data and uneven distribution among classes in machine learning tasks, a supervised generation method for called WAGAN cyclic sampling named SACS (Semi-supervised Active-learning Cyclic Sampling), based on semi-supervised active learning, are proposed. The loss function neural network structure optimized, quantity quality sample set enhanced. To enhance reliability generating pseudo-labels, Semi-supervised Active Framework (SAF) is...

10.3390/math12172634 article EN cc-by Mathematics 2024-08-24

With the development of computer technology, means human‐computer interaction are constantly updated. The trend is from human to adapt habits. In interaction, somatosensory sensor technology used track and analyze actions through vision technology. This paper adopts implement gesture recognition, which an important way in interaction. can obtain depth information bone First, decomposed into a sequence consisting micro gestures describe different directions. Then, compared with templates....

10.1002/itl2.311 article EN Internet Technology Letters 2021-07-10

Neuronal soma segmentation is essential for morphology quantification analysis. Rapid advances in light microscope imaging techniques have generated such massive amounts of data that time-consuming manual methods cannot meet requirements high throughput. However, touching still a challenge automatic methods. In this paper, we propose method combines the Rayburst sampling algorithm and ellipsoid fitting. The improved used to detect surface; fitting then refines jagged sampled surface generate...

10.1007/s12021-017-9336-y article EN cc-by Neuroinformatics 2017-09-22

Vega has been widely used in Virtual Reality (VR) field. infrared (IR) module can implement IR simulation, but imaging simulation's general approach does not apply to large-scale scene's simulation problem. This article deeps into scene image through based on visible image, the corresponding Digital Elevation Model (DEM) data and targets' 3D model. A real time system is successfully designed realized this paper. At aspect of classification, a coarse fine k-means clustering method consistency...

10.1109/m2rsm.2011.5697423 article EN 2011-01-01

Foreground segmentation in video frames is quite valuable for object and activity recognition, while the existing approaches often demand training data or initial annotation, which expensive inconvenient. We propose an automatic unsupervised method of foreground given unlabeled short video. The pixel-level optical flow binary mask features are converted into normal probabilistic superpixels, therefore, they adaptable to build superpixel-level conditional random field aims label background....

10.1117/1.jei.24.2.023034 article EN cc-by Journal of Electronic Imaging 2015-04-23

We consider the chance-constrained program (CCP) with random right-hand side under a finite discrete distribution. It is known that standard mixed integer linear programming (MILP) reformulation of CCP generally difficult to solve by general-purpose solvers as branch-and-cut search trees are enormously large, partly due weak relaxation. In this paper, we identify another reason for phenomenon: intersection feasible regions subproblems in tree could be nonempty, leading wasteful duplication...

10.48550/arxiv.2406.10472 preprint EN arXiv (Cornell University) 2024-06-14

Adverse health effects of Fe and S in airborne particulate matter (PM) have been reported. However, little work has done to characterize PM 10 from coal combustion. In this study, a sub-bituminous (coal A) bituminous B) were subjected combustion drop tube furnace under air-and oxy-firing conditions. Size distribution elemental composition (PM with aerodynamic diameter μm) obtained by low pressure impactor X-ray fluorescence techniques, respectively. The partitioning characteristics...

10.4028/www.scientific.net/amr.726-731.963 article EN Advanced materials research 2013-08-01

Through the error from feedback of classroom teaching system ,based on analysis and judgment intelligence system, can provide cause for students, automatically push corresponding after-school exercises , enhance learning relevance efficiency.

10.2991/ameii-15.2015.24 article EN cc-by-nc 2015-01-01

This paper analyzes the operational characteristics of library. Based on this, author improve traditional PageRank algorithm to rank for similar books. innovation is that according algorithm, reduced times, books, readers, book by three a unified relationship plane up, through matrix solution relations in initial value problems return, and achieve algorithm.

10.4028/www.scientific.net/amr.915-916.1266 article EN Advanced materials research 2014-04-09

10.1166/jctn.2014.3443 article EN Journal of Computational and Theoretical Nanoscience 2013-11-22

In order to study the various furnace arch structure and air distribution on combustion characteristics of chain boiler, based computational fluid dynamics theory, using numerical simulation method, a three-dimensional physical model for QXL1.4-0.7/95/70-AIItype hot water boiler is established, dividing grid, setting boundary conditions, then process flue gas flow were simulated, obtained temperature field concentration in furnace.The results show that increase length front decrease height...

10.2991/eame-17.2017.17 article EN cc-by-nc 2017-01-01

Cold rolling and recrystallization annealing processing of commercially pure titanium (CP-Ti) sheet was conducted. forming properties using a microstructural analysis, Vickers microhardness erichsen value measurements were investigated. The results have shown that alpha grain sizes 1 2 are 20.58μm 30.90μm, respectively. Erichsen test the specimens lowest values. This is because deformation reduction 75% increases crystal defects leads to high dislocation density. also became poor due 50%...

10.12783/dtetr/icace2018/25513 article EN DEStech Transactions on Engineering and Technology Research 2018-10-18
Coming Soon ...