A5101528191- Traffic Prediction and Management Techniques
- Optical measurement and interference techniques
- Human Mobility and Location-Based Analysis
- Time Series Analysis and Forecasting
- Advanced Measurement and Metrology Techniques
- Silicon Carbide Semiconductor Technologies
- Transportation Planning and Optimization
- Structural Health Monitoring Techniques
- Video Surveillance and Tracking Methods
- Image Processing Techniques and Applications
- Data Management and Algorithms
- High voltage insulation and dielectric phenomena
- Advancements in Semiconductor Devices and Circuit Design
- Random lasers and scattering media
- Advanced X-ray Imaging Techniques
- Advanced Measurement and Detection Methods
- Traffic control and management
- Optical Systems and Laser Technology
- Radar Systems and Signal Processing
- Optical Coherence Tomography Applications
- Semiconductor materials and devices
- Microwave Imaging and Scattering Analysis
- Air Quality Monitoring and Forecasting
- Human Pose and Action Recognition
- Graphene research and applications
Beijing University of Posts and Telecommunications
2019-2025
Southwest Jiaotong University
2025
University of Electronic Science and Technology of China
2025
Xi'an Polytechnic University
2025
Hangzhou Dianzi University
2021-2024
Nantong University
2024
Affiliated Hospital of Nantong University
2024
Southeast University
2012-2024
Southeast University
2019
Shandong Jiaotong University
2019
With the popularity of unmanned aerial vehicles (UAVs), how to conduct automatic and effective detection prevent unauthorized flying has become an important issue. The conventional constant false alarm rate (CFAR) detector based on radar signal shown advantages in moving target detection. However, CFAR-based detectors are strongly dependent some manual experience, such as ambient noise distribution estimation windows' size selection, usually suffered poor performance small UAV due weak...
Multiple works have applied deep learning to fringe projection profilometry (FPP) in recent years. However, obtain a large amount of data from actual systems for training is still tricky problem, and moreover, the network design optimization worth exploring. In this paper, we introduce graphic software build virtual FPP order generate desired datasets conveniently simply. The way constructing system described detail firstly, then some key factors set much closer reality are analyzed. With...
Forecasting traffic flow and speed in the urban is important for many applications, ranging from intelligent navigation of map applications to congestion relief city management systems. Therefore, mining complex spatio-temporal correlations data accurately predict essential community. However, previous studies that combined graph convolution network or self-attention mechanism with deep time series models (e.g., recurrent neural network) can only capture spatial dependencies each slot...
Traffic forecasting is crucial for public safety and resource optimization, yet very challenging due to the temporal changes dynamic spatial correlations of traffic data. To capture these intricate dependencies, spatio-temporal networks, such as recurrent neural networks with graph convolution attention full are applied. However, previous based on end-to-end training thus fail handle distribution shift in non-stationary time series. On other hand, efficient effective algorithm modeling still...
In recent years, with the rapid development of public transportation, ways people travel has become more diversified and complicated. Transportation mode detection, as a significant branch human activity recognition (HAR), is great importance in analyzing patterns, traffic prediction planning. Though many works have been devoted to transportation there remains challenge for accurate robust pattern identification. this paper, we propose residual LSTM recurrent networks-based detection...
Harvesting the energy from interaction between hygroscopic materials and atmospheric water can generate green clean energy. However, ion diffusion process of moisture-induced dissociation leads to disappearance concentration gradient gradually, there is still a lack moisture-based power generation devices with truly continuous operation, especially duration current output needs be extended. Here, we propose design for reconstructing by coupling photocatalytic hydrogen evolution reaction...
To enhance the electrocatalytic performance of a flexible Pd@CFs catalyst for methanol oxidation, deep cryogenic treatment in liquid nitrogen was introduced. The effects frequency and time on surface crystal orientation, microstructure morphology, mechanical performance, oxidation were studied. results showed that when 2 times 24 h, best. Compared with without treatment, activity stability increased by about 33% 41%, respectively. 43.4 6.3 commercial Pd/C catalyst, After 500 cycles CV...
Travel Time Estimation (TTE) is indispensable in intelligent transportation system (ITS). It significant to achieve the fine-grained Trajectory-based (TTTE) for multi-city scenarios, namely accurately estimate travel time of given trajectory multiple city scenarios. However, it faces great challenges due complex factors including dynamic temporal dependencies and spatial dependencies. To tackle these challenges, we propose a meta learning based framework, MetaTTE, continuously provide...
Travel time estimation(TTE) is a critical component of intelligent transportation systems. To achieve efficient and accurate trajectory-based travel estimation, it essential to design streamlined model that reduces computation memory costs. However, this challenging as traditional deep neural networks are limited in their calculation capabilities can be cumbersome due the high number parameters. overcome these challenges, we propose novel approach utilizing well-designed network called...
With the rapid development of mobile Internet techniques, using sensor-rich smartphones to sense various contexts attracts much attention, such as transportation mode recognition. The information can help improve urban planning, traffic management and journey planning. Though work has been done on recognition classic machine learning algorithms, performance these methods is not reasonable heavily relies effectiveness handcrafted features. In this paper, we leverage strong representation...
Abstract Central nervous system (CNS) diseases encompass spinal cord injuries, brain tumors, neurodegenerative diseases, and ischemic strokes. Recently, there has been a growing global recognition of CNS disorders as leading cause disability death in humans the second most common worldwide. The burdens treatment challenges posed by are particularly significant context rapidly expanding population aging demographics. blood-brain barrier (BBB) presents challenge for effective drug delivery...
Mitophagy influences the progression and prognosis of ischemic stroke (IS). However, whether DNA methylation in brain is associated with altered mitophagy hypoxia-injured neurons remains unclear. Here, miR-138–5p was found to be highly expressed exosomes secreted by astrocytes stimulated oxygen glucose deprivation/re-oxygenation (OGD/R), which could influence recovery OGD/R-injured through autophagy. Mechanistically, promotes stable expression Ras homolog enriched like 1(Rhebl1)...
Traffic forecasting belongs to intelligent transportation systems and is helpful for public property life safety. Therefore, forecast traffic accurately, researchers pay great attention dealing with complex problems by mining intricate spatial temporal dependencies of the traffic. However, some challenges still hold back forecasting: 1) Most studies mainly focus on modeling correlations time series close distances road network ignore remote but similar series; 2) Previous static graph-based...
Fringe-based optical measurement techniques require reliable fringe analysis methods, where empirical mode decomposition (EMD) is an outstanding one due to its ability of analyzing complex signals and the merit being data-driven. However, two challenging issues hinder application EMD in practical measurement. One tricky mixing problem (MMP), making decomposed intrinsic functions (IMFs) have equivocal physical meaning; other automatic accurate extraction sinusoidal from IMFs when...
Next Point-of-Interest (POI) recommendation is a pivotal issue for researchers in the field of location-based social networks. While many recent efforts show effectiveness recurrent neural network-based next POI algorithms, several important challenges have not been well addressed yet: (i) The majority previous models only consider dependence consecutive visits, while ignoring intricate dependencies POIs traces; (ii) nature hierarchical and matching sub-sequence sequences are hardly model...
The effectiveness of the renowned empirical mode decomposition (EMD) is affected by mode-mixing problem (MMP) if a signal contains intermittent modes. ensemble EMD (EEMD) and several modified extended algorithms solve this adding random white noises. However, necessary large size inevitable manual intervention limits application EEMD. In letter, novel regenerated phase-shifted sinusoid-assisted (RPSEMD) proposed. Sinusoids with different scales are iteratively generated added to cope all...
In recent years, traffic forecasting has gradually attracted attention in data mining because of the increasing availability large-scale data. However, it faces substantial challenges complex temporal-spatial correlations traffic. Recent studies mainly focus on modeling local spatial by utilizing graph neural networks and neglect influence long-distance correlations. Besides, most existing works utilize recurrent networks-based encoder–decoder architecture to forecast multistep volume suffer...
The phase demodulation method of adaptive windowed Fourier transform (AWFT) is proposed based on Hilbert-Huang (HHT). HHT analyzed and performed fringe pattern to obtain instantaneous frequencies firstly. These are further the condition AWFT locate local stationary areas where fundamental spectrum will not be interfered by high-order spectrum. Within each area, can extracted accurately adaptively using with background, which has been determined previously presented criterion during HHT,...