A5052480553- Traffic Prediction and Management Techniques
- Traffic control and management
- Transportation Planning and Optimization
- Video Surveillance and Tracking Methods
- Remote-Sensing Image Classification
- Anomaly Detection Techniques and Applications
- Advanced Neural Network Applications
- Complex Network Analysis Techniques
- Text and Document Classification Technologies
- X-ray Diffraction in Crystallography
- Crystallization and Solubility Studies
- Advanced Clustering Algorithms Research
- Automated Road and Building Extraction
- Remote Sensing and LiDAR Applications
- Autonomous Vehicle Technology and Safety
- Fire Detection and Safety Systems
- Laser-induced spectroscopy and plasma
- Remote Sensing and Land Use
- Face and Expression Recognition
- Geochemistry and Geologic Mapping
- Time Series Analysis and Forecasting
- Robotic Path Planning Algorithms
- Human Mobility and Location-Based Analysis
- Analytical chemistry methods development
- Vehicle License Plate Recognition
East China Jiaotong University
2016-2025
South China Agricultural University
2024
Southwest Jiaotong University
2024
China University of Geosciences
2020-2024
Xinjiang University
2024
APC Microbiome Institute
2024
Hong Kong Polytechnic University
2024
East China University of Political Science and Law
2023
University of Florida
1985-2022
Chinese Academy of Sciences
2005-2021
Deep learning has achieved great successes in conventional computer vision tasks. In this paper, we exploit deep techniques to address the hyperspectral image classification problem. contrast tasks that only examine spatial context, our proposed method can both context and spectral correlation enhance classification. particular, advocate four new models, namely, 2-D convolutional neural network (2-D-CNN), 3-D-CNN, recurrent CNN (R-2-D-CNN), 3-D-CNN (R-3-D-CNN) for We conducted rigorous...
Road information extraction based on aerial images is a critical task for many applications, and it has attracted considerable attention from researchers in the field of remote sensing. The problem mainly composed two subtasks, namely, road detection centerline extraction. Most previous studies rely multistage-based learning methods to solve problem. However, these approaches may suffer well-known propagation errors. In this paper, we propose novel deep model, recurrent convolution neural...
With the recent explosive growth of e-commerce and online communication, a new genre text, short has been extensively applied in many areas. So researches focus on text mining. It is challenge to classify owing its natural characters, such as sparseness, large-scale, immediacy, non-standardization. difficult for traditional methods deal with classification mainly because too limited words cannot represent feature space relationship between documents. Several reviews are shown times. However,...
Kmeans-type clustering aims at partitioning a data set into clusters such that the objects in cluster are compact and different well separated. However, most kmeans-type algorithms rely on only intracluster compactness while overlooking intercluster separation. In this paper, series of new by extending existing is proposed integrating both First, objective functions for developed. Based these functions, corresponding updating rules then derived analytically. The properties performances...
Camera-based systems are increasingly used for collecting information on intersections and arterials. Unlike loop controllers that can generally be only detection movement of vehicles, cameras provide rich about the traffic behavior. Vision-based frameworks multiple-object detection, object tracking, near-miss have been developed to derive this information. However, much work currently addresses processing videos offline. In article, we propose an integrated two-stream convolutional networks...
Abstract Ultrasound is an acoustic wave which can noninvasively penetrate the skull to deep brain regions, enabling neuromodulation. However, conventional ultrasound’s spatial resolution diffraction-limited and low-precision. Here, we report nanobubble-mediated ultrasound stimulation capable of localizing effects only desired region in male mice. By varying delivery site nanobubbles, could activate specific regions mouse motor cortex, evoking EMG signaling limb movement, also, separately,...
Hyperspectral image (HSI) classification is an essential task in remote sensing with substantial practical significance. However, most existing convolutional neural network (CNN)-based methods focus only on local spatial features while neglecting global spectral dependencies. Meanwhile, Transformer-based exhibit robust capabilities for feature modeling but struggle to extract effectively. To fully exploit the extraction of CNN-based networks and networks, this paper proposes a dual-branch...
Accurate hyperspectral image classification has been an important yet challenging task for years. With the recent success of deep learning in various tasks, 2-dimensional (2D)/3-dimensional (3D) convolutional neural networks (CNNs) have exploited to capture spectral or spatial information images. On other hand, few approaches make use both and simultaneously, which is critical accurate classification. This paper presents a novel Synergistic Convolutional Neural Network (SyCNN) The SyCNN...
de niveau recherche, publiés ou non, émanant des établissements d'enseignement et recherche français étrangers, laboratoires publics privés.
Geological remote sensing interpretation (GRSI), which aims to recognize multiple geological elements based on their characteristics satellite images, is vital in large-scale regional lithological mapping. However, due the influence of long-term movements, spatial distribution (such as lithology, glaciers, and soils) image often complex highly fragmented. In addition, high inter-class similarity severe homogenization make annotation element samples require significant cost expert knowledge....
Road extraction is a crucial aspect of remote sensing imagery processing that plays significant role in various applications, including automatic driving, urban planning, and path navigation. However, accurate road challenging task due to factors such as high density, building occlusion, complex traffic environments. In this study, Spatial Attention Swin Transformer (SASwin Transformer) architecture proposed create robust encoder capable extracting roads from imagery. architecture, we have...
A commercial butane micron troch was used to enhance plasma optical emissions in laser-induced breakdown spectroscopy (LIBS). Fast imaging and spectroscopic analyses were observe evolution the atmospheric pressure for LIBS without with using a micro torch. Optical emission intensities signal-to-noise ratios (SNRs) as functions of delay time studied. Enhanced SNRs obtained by The effects laser pulse energy on same spectral intensity could be torch much lower energy. investigation SNR at...
Despite much effort and significant progress in recent years, image segmentation remains a challenging problem processing, especially for the low contrast, noisy synthetic aperture radar (SAR) images. This paper explores of oil slicks using partial differential equation (PDE)‐based level set method, which represents slick surface as an implicit propagation interface. Starting from initial estimation with priori information, method creates speed functions to detect position Specifically,...
The spatial-temporal prediction of traffic flow is very important for management and planning. most difficult challenges are the temporal feature extraction spatial correlation nodes. Due to complex between different roads dynamic trend time patterns, traditional forecasting methods still have limitations in obtaining correlation, which makes it extract more valid information. In order improve accuracy forecasting, this paper proposes a multi-scale dual graph convolution network (MD-GCN)....
The objective of this paper is threefold. First, the feasibility modeling a large-scale network at microscopic level detail presented. Second, unique data collection challenges that are involved in constructing and calibrating microscopically described. Third, opportunities applications from use as opposed to macroscopic simulation tool possibility using model demonstrated. requirements validated for are: ( a) must be capable origin-destination demand tables, b) dynamic traffic routing, c)...