A5102882932- Pulsars and Gravitational Waves Research
- Gamma-ray bursts and supernovae
- Advanced Algorithms and Applications
- Advanced Sensor and Control Systems
- Cloud Computing and Resource Management
- Laser-Matter Interactions and Applications
- Energy Efficient Wireless Sensor Networks
- Welding Techniques and Residual Stresses
- Industrial Vision Systems and Defect Detection
- Machine Fault Diagnosis Techniques
- Human Pose and Action Recognition
- Nuclear physics research studies
- Advanced Vision and Imaging
- Non-Destructive Testing Techniques
- Network Time Synchronization Technologies
- Innovative Educational Techniques
- Distributed and Parallel Computing Systems
- Solar-Powered Water Purification Methods
- Vehicle License Plate Recognition
- Video Surveillance and Tracking Methods
- Network Traffic and Congestion Control
- Advanced Chemical Physics Studies
- Gear and Bearing Dynamics Analysis
- IoT and Edge/Fog Computing
- Advanced Neural Network Applications
Taiyuan Normal University
2025
University of Hong Kong
2024
Guangxi University
2019-2024
Hong Kong University of Science and Technology
2024
Hunan University of Traditional Chinese Medicine
2024
Wuhan University of Technology
2024
University of Electronic Science and Technology of China
2023
Dalian Maritime University
2023
Stomatology Hospital
2021
Kunming Medical University
2021
Temporal grounding aims to localize temporal boundaries within untrimmed videos by language queries, but it faces the challenge of two types inevitable human uncertainties: query uncertainty and label uncertainty. The uncertainties stem from subjectivity, leading limited generalization ability grounding. In this work, we propose a novel DeNet (Decoupling Debias) embrace uncertainty: Decoupling — We explicitly disentangle each into relation feature modified feature. feature, which is mainly...
Previous deep learning-based brain network research has made significant progress in understanding the pathophysiology of schizophrenia. However, it ignores three-dimensional spatial characteristics EEG signals and cannot dynamically learn interactions between nodes. To address this issue, a schizophrenia classification model based on adaptive graph convolutional neural (3D-AGCN) is proposed. Each subject's data divided into various segment lengths frequency bands for experiment. The...
Abstract Effective compression of densely sampled BRDF measurements is critical for many graphical or vision applications. In this paper, we present DeepBRDF, a deep‐learning‐based representation that can significantly reduce the dimensionality measured BRDFs while enjoying high quality recovery. We consider each as sequence image slices and design deep autoencoder with masked L 2 loss to discover nonlinear low‐dimensional latent space high‐dimensional input data. Thorough experiments verify...
We investigate high-order harmonic generation (HHG) from asymmetric molecules exposed to intense laser fields. show that the emissions of odd and even harmonics depend differently on orientation angle, internuclear distance, as well effective charge. This difference mainly comes different roles intramolecular interference in HHG harmonics. These map structure molecule vs spectra.
We investigate the emission times of high-order harmonics from asymmetric molecule HeH${}^{2+}$ exposed to intense laser fields numerically and analytically. show that for each cycle, follow two characteristic electron trajectories. The first one matches prediction general ground-state-based quantum orbital theory (GSB-QOT), as second differs significantly GSB-QOT. excited state plays an important role in this difference. As a result, can be used tool monitor dynamics states asymmetry strong fields.
In the applications of solar magneto-seismology(SMS), employing ratio period fundamental mode to twice one its first overtone, $P_1/2P_2$, plays an important role. We examine how field-aligned flows affect dispersion properties, and hence ratios, standing modes supported by magnetic slabs in atmosphere. numerically solve relations devise a graphic means construct modes. For coronal slabs, we find that flow effects are significant, for fast kink sausage alike. ones, they may reduce $P_1/2P_2$...
A liquid pump is the core component of a cooling system. With increase in heat dissipation airborne electronic equipment, conventional air-cooled motor can no longer meet requirements. To system requirements, certain overloaded, which puts forward higher requirements for cooling. effect motor, this paper takes and its as research object. It four structural schemes circuit according to different arrangement modes bends, through holes, fins circuit. The whole flow field modelled three...
The attosecond technology provides a powerful tool for studying the ultrafast dynamics of electrons during strong-field ionization atoms and molecules. This relies on quantitative theoretical models to invert time-domain information system process from photoelectron spectrum obtained through experimental measurements. One key issues in constructing is description Coulomb effect. Compared with single-center potential atoms, molecules have multi-center coulomb potential. fundamental geometric...
Waste tires have excellent mechanical performance and been used as reinforcing material in geotechnical engineering; however, their interface properties are poorly understood. To further our knowledge, this paper examines the pull-out characteristics of waste tire strips a compacted sand, together with uniaxial biaxial geogrids also tested under same conditions. The analysis results shows that interlocking effect resistance between strip sand is very strong significantly higher than...
When a high-speed train is running, it easy for foreign objects like rail-side plastic bags to enter bottom bogies, cables and equipment gaps, which affects the safety of driving. At present, detection accuracy such low. To solve this problem, present study used latest deep learning based object networks, as SSD Faster R-CNN, combined with different feature extractors build network. Through data augmentation sample, number sizes was expanded. By idea transfer learning, COCO dataset source...