A5012473461- 3D Shape Modeling and Analysis
- Computer Graphics and Visualization Techniques
- Gold and Silver Nanoparticles Synthesis and Applications
- Advanced Numerical Analysis Techniques
- Cancer Cells and Metastasis
- Electrochemical Analysis and Applications
- Advanced Nanomaterials in Catalysis
- Advanced Neural Network Applications
- Computational Geometry and Mesh Generation
- Generative Adversarial Networks and Image Synthesis
- Liver Disease Diagnosis and Treatment
- Cancer, Stress, Anesthesia, and Immune Response
- Advanced Vision and Imaging
- Adversarial Robustness in Machine Learning
- Electrocatalysts for Energy Conversion
- Cancer Immunotherapy and Biomarkers
- Dye analysis and toxicity
- Hand Gesture Recognition Systems
- Membrane Separation and Gas Transport
- 3D Surveying and Cultural Heritage
- Machine Learning and ELM
- Forecasting Techniques and Applications
- Neural Networks and Applications
- Teleoperation and Haptic Systems
- Stock Market Forecasting Methods
Chinese University of Hong Kong, Shenzhen
2023-2024
Institute of Biophysics
2024
Chinese Academy of Sciences
2024
University of Chinese Academy of Sciences
2024
Xi'an Technological University
2020-2024
Tongji Hospital
2024
Huazhong University of Science and Technology
2023-2024
Boston University
2021-2024
Texas A&M University
2024
Shenyang Institute of Automation
2024
Geometric deep learning has sparked a rising interest in computer graphics to perform shape understanding tasks, such as classification and semantic segmentation. When the input is polygonal surface, one suffer from irregular mesh structure. Motivated by geometric spectral theory, we introduce <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Laplacian2Mesh</i> , novel flexible convolutional neural network (CNN) framework for coping with...
The task of explicit surface reconstruction is to generate a mesh by interpolating given point cloud. Explicit necessary when the cloud required appear exactly on surface. However, for non-perfect input, such as lack normals, low density, irregular distribution, thin and tiny parts, high genus, robust method that can high-quality manifold triangulation missing. We propose starts from an initial simple mesh, alternately performs Filmsticking step Sculpting converges interpolates all input...
Machine Learning (ML) techniques have been applied to many real-world applications perform a wide range of tasks. In practice, ML models are typically deployed as the black-box APIs protect model owner's benefits and/or defend against various privacy attacks. this paper, we present Gradient-Leaks first evidence showcasing possibility performing membership inference attacks (MIAs), with mere access, which aim determine whether data record was utilized train given target or not. The key idea...
Anaplastic lymphoma kinase (ALK) is a highly responsive therapeutic target for
Purpose To assess the utility of fat fraction quantification using quantitative multi-echo Dixon for evaluating tumor proliferation and microvascular invasion (MVI) in hepatocellular carcinoma (HCC). Methods A total 66 patients with resection histopathologic confirmed HCC were enrolled. Preoperative MRI proton density R 2 * mapping was analyzed. Intratumoral peritumoral regions delineated manually placed interest at maximum level intratumoral fat. Correlation analysis explored relationship...
A method for detecting relations between symbols in handwritten mathematical expressions is presented. Expressions are automatically decomposed into blocks by projections onto x or y axes. The represented rectangles surrounding symbols. authors discuss aligned which exist. These categorized as being on the same line, exponents subscripts. statistical labeling associates probability of each relation with block pair. Final obtained propagating local information using contextual constraints....
Feature lines are important geometric cues in characterizing the structure of a CAD model. Despite great progress both explicit reconstruction and implicit reconstruction, it remains challenging task to reconstruct polygonal surface equipped with feature lines, especially when input point cloud is noisy lacks faithful normal vectors. In this paper, we develop multistage algorithm, named RFEPS, address challenge. The key steps include (1)denoising based on assumption local planarity,...
The rapid expansion of the non-fungible token (NFT) market has attracted many investors. However, studies on NFT price fluctuations have been relatively limited. To date, machine learning approach not used to demonstrate a specific error in sale fluctuation prediction. aim this study was develop prediction model for using trading information obtained from OpenSea, world’s largest marketplace. We Python programs collect data and summarized them as: information, collection related account...
Surface reconstruction is a challenging task when input point clouds, especially real scans, are noisy and lack normals. Observing that the Multilayer Perceptron (MLP) implicit moving least-square function (IMLS) provide dual representation of underlying surface, we introduce <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Neural-IMLS</i> , novel approach directly learns noise-resistant signed distance (SDF) from unoriented raw clouds in...
The development of electrocatalysts is vital importance to the OER. In this work, an optimized MOF electrode, Ir-NiFe-btz/NF, was synthesized by incorporating ultralow levels Ir nanoparticles into ionic bimetallic MOF. Under alkaline conditions, prepared electrode demonstrated outstanding OER activity and stability with a low overpotential 182 mV reach 10 mA cm
Neural implicit representation is a promising approach for reconstructing surfaces from point clouds. Existing methods combine various regularization terms, such as the Eikonal and Laplacian energy to enforce learned neural function possess properties of Signed Distance Function (SDF). However, inferring actual topology geometry underlying surface poor-quality unoriented clouds remains challenging. In accordance with Differential Geometry, Hessian SDF singular points within differential...
Deep learning (DL) has achieved tremendous success in recent years many fields. The of DL typically relies on a considerable amount training data and the expensive model optimization process. Therefore, trained its corresponding have become valuable assets whose intellectual property (IP) needs to be protected. Once or dataset is released, there currently no mechanism for entity that owns one part establish clear relationship with other. In this paper, we aim reveal integrated between given...
Abstract Metal–organic frameworks (MOFs) are promising electrocatalysts for clean energy conversion systems. However, developing MOF‐based electrodes with high performance toward oxygen evolution reaction (OER) is still challenging. In this work, a series of MOF film derived from Ni‐btz were prepared by employing the secondary growth strategy under solvothermal conditions. Fe and Co ions also incorporated into framework to produce trimetallic coupling effect obtain enhanced OER activity. The...
Abstract This article analyzes the spread and infiltration dynamic process of liquid droplets on surface porous materials establishes a mathematical model changes in droplets. To accurately describe effects droplet flow, dynamics was developed by applying level-set method, seepage also analyzed numerically. The series control parameters percolation are analyzed, pore degree performed numerical study deformation process. We observe models. It found that there is competitive relationship...