A5029865350- Geophysical and Geoelectrical Methods
- Seismic Imaging and Inversion Techniques
- Geophysics and Gravity Measurements
- Geophysical Methods and Applications
- Seismic Waves and Analysis
- Ombudsman and Human Rights
- Earthquake Detection and Analysis
- Geomagnetism and Paleomagnetism Studies
- Law and Political Science
- Heat Transfer and Optimization
- Digital Marketing and Social Media
- Electrostatics and Colloid Interactions
- Spectroscopy and Quantum Chemical Studies
- Complex Systems and Time Series Analysis
- Scoliosis diagnosis and treatment
- Engineering Diagnostics and Reliability
- Advanced Measurement and Metrology Techniques
- Matrix Theory and Algorithms
- Heat transfer and supercritical fluids
- Expert finding and Q&A systems
- Heat Transfer and Boiling Studies
- Risk and Safety Analysis
- Recommender Systems and Techniques
- Geological Modeling and Analysis
- Emotion and Mood Recognition
Jilin University
2015-2025
Jilin Province Science and Technology Department
2010-2025
Civil Aviation University of China
2025
Jilin Medical University
2024
Southeast University
2022-2023
State Key Laboratory of Digital Medical Engineering
2022-2023
China University of Geosciences
2023
University of Hong Kong
2023
Zhejiang Normal University
2023
Taiyuan University of Technology
2022
A novel heterojunction composite of CoOx/Bi4Ti3O12 was synthesized through a combination molten salt and photodeposition methods. The optimal sample exhibited superior performance in the piezocatalytic degradation methyl orange (MO) dye with rate 1.09 h−1, which 2.4 times higher than that pristine Bi4Ti3O12. Various characterizations were conducted to reveal fundamental nature accountable for outstanding CoOx/Bi4Ti3O12. investigation band structure indicated formed type-I p-n structure, CoOx...
Dimensionality reduction (DR) is important for feature extraction and classification of hyperspectral images (HSIs). Recently proposed superpixel-based DR models have shown promising performance, where superpixel segmentation techniques were applied to segment an HSI then like principal component analysis (PCA) or linear discriminant (LDA) employed extract the local and/or global features. However, superpixelwise PCA based features are unsatisfactory because aims with high variance, which...
This paper proposes a physics-constrained 3D Swin Transformer (ST) for gravity inversion. By leveraging the self-attention mechanism in ST, method effectively models global dependencies within data, enabling network to reweight features globally and focus on critical anomalous regions. Additionally, prior gradient information is integrated into loss function, hierarchical weight allocation strategy adopted guide model learning boundary of density structures deep-seated more effectively....
Infrared small target detection (IRSTD) aims to segment targets from an infrared clutter background. However, the long imaging distance, complex background, and extremely limited number of pixels pose great challenges for IRSTD. In this paper, we propose a new IRSTD method based on deep graph neural network fully extract fuse texture structural information images. Firstly, clustering algorithm is designed divide image into several subgraphs as prior knowledge guide initialization structure...
As a key component of remote-sensing technology, satellite gravity observation offers extensive coverage and high accuracy, effectively compensating for the shortcomings terrestrial measurements. Three-dimensional data inversion can predict physical property spatial distribution geological formations beneath surface by analyzing data. In this paper, heat source position within Gonghe Basin’s geothermal system is identified through analysis data, constrained deep-learning method proposed....
The nonlinear model predictive control (NMPC) controller is designed for an engine cooling system and aims to the pump speed fan according thermal load, vehicle speed, ambient temperature in real time with respect coolant comprehensive energy consumption of system, which serve as targets. connected underhood computational fluid dynamics (CFD) by coupling transmission equation. For intricate management process performance analysis, a multi-thermodynamic integrated was established. concept...
Three-dimensional gravity inversion is an effective method for restoring underground density distribution from anomaly data. Conventional regularization has good data fitting, but its model insufficient fitting capabilities due to low-depth resolution. Although data-driven deep learning-based results significantly improve depth resolution and physical property distribution, it difficult ensure the of results. Accordingly, this study proposes a three-dimensional based on enhanced dual U-Net...
The Cobb angle (CA) serves as the principal method for assessing spinal deformity, but manual measurements of CA are time-consuming and susceptible to inter- intra-observer variability. While learning-based methods, such SpineHRNet+, have demonstrated potential in automating measurement, their accuracy can be influenced by severity image quality, relative position rib vertebrae, etc. Our aim is create a reliable approach that provides consistent highly accurate from posteroanterior (PA)...
This study reports on the characterization of reporter dye-loaded block copolymer vesicles (polymersomes) PS115-b-PAA15 (polystyrene-block-poly(acrylic acid)) and PEG114-b-PLA167 (poly(ethylene glycol)-block-poly(lactic their drying behavior by fluorescence lifetime imaging microscopy (FLIM). The characteristic changes decay components dye calcein incorporated in three different local nanoenvironments, namely, solvated dye, associated with vesicle wall, dried agglomerated are observed FLIM...
Collaborative filtering is key technique of recommendation system. But traditional collaborative methods are inefficient especially when the user-rating data extremely sparse. To solve this problem, we propose an approach to compute user similarity with type users-rating items in paper, and then develop a algorithm based on approach. Furthermore, put forward improved combination, which combines types items. Last, carry out experiment classic Movielens sets evaluate algorithm, use MAE as...
Electroencephalogram (EEG) signal is often used in emotion recognition tasks to classify human emotions. In this paper, we propose a new approach learn the temporal features of EEG using long and short-term memory (LSTM), which type Recurrent Neural network (RNN), especially suitable for solving problem long-term dependencies such as gradients vanishing exploding. addition, enhance interaction between signals non-linear characteristics electrodes, use 1D-Convolution kernel pre-process input...
Three-dimensional magnetic inversion allows the distribution of parameters to be obtained, and it is an important tool for geological exploration interpretation. However, because redundancy data obtained from large-scale investigations or high-density sampling, very computationally intensive use these iterative calculations. In this paper, we propose a method compressing by using adaptive quadtree decomposition method, which divides two-dimensional region into four quadrants progressively...
Following the experience economy era coming, in order to retain customers, we must focus on user and so do banking industry. Experience is core engine of network financial. In this paper, study e-banking service improvement with breakthrough point experience. This paper bases technology acceptance model refer multiple extension at same time, build verify electronic bank adoption model. It concluded factors influencing services. Then guide design based theory contact that improve experience,...
Gravity inversion can be used to obtain the spatial structure and physical properties of subsurface anomalies through gravity observation data. With continuous development machine learning, geophysical methods based on deep learning have achieved good results. Geophysical often employ large-scale data sets networks with strong generalization. They are widely but face a problem lacking information constraints. Therefore, self-constrained network is proposed optimize results, composed two...