A5101433125- Advanced Vision and Imaging
- Vasculitis and related conditions
- Reinforcement Learning in Robotics
- Renal Diseases and Glomerulopathies
- Species Distribution and Climate Change
- Image Processing Techniques and Applications
- 3D Shape Modeling and Analysis
- Heart Failure Treatment and Management
- Robotics and Sensor-Based Localization
- Robotic Mechanisms and Dynamics
- 3D Surveying and Cultural Heritage
- Artificial Intelligence in Healthcare
- Sarcoidosis and Beryllium Toxicity Research
- Remote Sensing and LiDAR Applications
- Machine Learning in Healthcare
- Data Visualization and Analytics
- Computer Graphics and Visualization Techniques
- Robot Manipulation and Learning
- Artificial Immune Systems Applications
- Image Retrieval and Classification Techniques
- Subcritical and Supercritical Water Processes
- Photonic Crystal and Fiber Optics
- Humor Studies and Applications
- Generative Adversarial Networks and Image Synthesis
- Cardiac Fibrosis and Remodeling
Zhejiang University
2009-2024
Shanghai University of Traditional Chinese Medicine
2024
First Affiliated Hospital Zhejiang University
2021-2023
Stanford University
2018-2022
Southeast University
2022
Jiangsu Industry Technology Research Institute
2022
Tabor College
2022
Zhejiang Center for Disease Control and Prevention
2021
Tsinghua University
2020
Center for Information Technology
2020
Previous studies in multimodal sentiment analysis have used limited datasets, which only contain unified annotations. However, the annotations do not always reflect independent of single modalities and limit model to capture difference between modalities. In this paper, we introduce a Chinese single- multi-modal dataset, CH-SIMS, contains 2,281 refined video segments wild with both unimodal It allows researchers study interaction or use for analysis.Furthermore, propose multi-task learning...
We demonstrate model-based, visual robot manipulation of deformable linear objects. Our approach is based on a state-space representation the physical system that aims to control. This choice has multiple advantages, including ease incorporating physics priors in dynamics model and perception model, planning actions. In addition, states can naturally represent object instances different appearances. Therefore, state space be learned one setting directly used other visually settings. contrast...
Estimating crop yield in large areas is essential for ensuring food security and sustainable development. Accounting variations the temporal cumulative growth of crops across regions (i.e., spatial heterogeneity growth) can improve accuracy estimation areas. However, current learning methods have limitations such as cutting off inherent correlations among regions, difficulty obtaining accurate prior knowledge, high subjectivity. Therefore, this study proposed a novel deep adaptive model...
Spatial downscaling is an important approach to obtain high-resolution land surface temperature (LST) for thermal environment research. However, existing methods are unable sufficiently address both spatial heterogeneity and complex nonlinearity, especially in scenes (<120 m), accordingly limit the representation of regional details accuracy inversion. In this study, by integrating normalized difference vegetation index (NDVI), building (NDBI), digital elevation model (DEM), slope data, a...
Abstract We propose a novel approach to learning cloth deformation as function of body pose, recasting the graph‐like triangle mesh data structure into image‐based in order leverage popular and well‐developed convolutional neural networks (CNNs) two‐dimensional Euclidean domain. Then, three‐dimensional animation clothing is equivalent sequence RGB images driven/choreographed by time dependent joint angles. In reduce nonlinearity demands on network, we utilize procedural skinning surface...
With the aim of creating virtual cloth deformations more similar to real world clothing, we propose a new computational framework that recasts three dimensional deformation as an RGB image in two pattern space. Then animation is equivalent sequence images, which turn are driven/choreographed via parameters such joint angles. This allows us leverage popular CNNs learn The pixels extended into standard body skinning techniques, after values interpreted texture offsets and displacement maps....
Objective We aimed to validate and modify the renal risk score for antineutrophil cytoplasmic antibody (ANCA)-associated glomerulonephritis (AAGN) in a Chinese cohort with majority of myeloperoxidase (MPO)-positive patients. Methods A total 285 patients biopsy-proven AAGN our center were retrospectively included. Patients randomly assigned development set (n = 201) validation 84). calculated analyzed clinicopathological characteristics follow-up data. The nomogram was constructed based on...
In this paper, we approach the challenging problem of motion planning for knot tying. We propose a hierarchical in which top layer produces topological plan and bottom translates into continuous robot motion. The decomposes knotting task sequences abstract actions based on theory. each these trajectories through learned primitives. To adapt action to specific rope geometry, primitives take observed configuration as input. train by imitating human demonstrations reinforcement learning...
Dry weight (DW), defined as the lowest tolerated postdialysis following ultrafiltration (UF) of excess fluid volume, is essential for any dialysis prescription hemodialysis (HD) patients. However, there no gold standard DW assessment, and difficulty its accurate assessment increases given individual variations dynamic changes caused by uncertainty patients' condition. Therefore, current empirical evaluation process often crude, imprecise, experience-dependent, energy-consuming. Here, we...
The technical feasibility of using calcium oxide (CaO) as a sorbent for CO2 and Fe-Cr catalyst the water-gas shift (WGS) reaction syngas steam gasification biomass was investigated.The effects temperature, to mass ratio, CaO molar WGS on gas composition were studied.Within temperature range 250 C 550 C, H2 concentration increased from 1.2% 17.1%, with total increase 16%.As rate within 0 kg/h 0.12 kg/h, maximum value 12.1% 17.13%, 5%.As ratio 2, demonstrated minimum 1.3%, exhibited 53.1%.A...
Background Kidney involvement is common in antineutrophil cytoplasmic antibody (ANCA) associated vasculitis (AAV). It tends to be aggressive, and some patients, the kidney may reach criteria of acute injury (AKI). Here, we aim describe clinical characteristics these patients find risk factors for poor outcomes. Methods Patients diagnosed with AAV our hospital from February 2003 2017 were included. Those who reached KDIGO AKI reclassified according stage. The features analyzed. Also,...
Abstract Introduction Pretransplant osteoporosis and vascular calcification probably increase the risk of fractures cardiovascular events after kidney transplantation. In present study, we investigated related factors among end-stage renal disease (ESRD) patients awaiting Methods A total 221 ESRD (age, 43.4 ± 14.3 years; 125 males 96 females; median dialysis duration, 61.0 m) transplantation were enrolled in this cross-sectional study. Serum levels bone turnover markers intact parathyroid...
In high-end visual effects pipelines, a customized (and expensive) light stage system is (typically) used to scan an actor in order acquire both geometry and texture for various expressions. Aiming towards democratization, we propose novel pipeline obtaining as well enough expression information build person-specific animation rig without using or any other hardware (or manual cleanup). A key idea consists of warping real-world images align with the template avatar subsequently projecting...
We tackle the challenging problem of creating full and accurate three dimensional reconstructions botanical trees with topological geometric accuracy required for subsequent physical simulation, e.g. in response to wind forces. Although certain aspects our approach would benefit from various improvements, results exceed state art especially complexity accuracy. Starting two RGB image data acquired cameras attached drones, we create point clouds, textured triangle meshes, a simulatable...
We demonstrate model-based, visual robot manipulation of linear deformable objects. Our approach is based on a state-space representation the physical system that aims to control. This choice has multiple advantages, including ease incorporating physics priors in dynamics model and perception model, planning actions. In addition, states can naturally represent object instances different appearances. Therefore, state space be learned one setting directly used other visually settings. contrast...