Yu Hu

ORCID: 0000-0001-8006-7659
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Electronic Packaging and Soldering Technologies
  • Welding Techniques and Residual Stresses
  • Hydrogen embrittlement and corrosion behaviors in metals
  • 3D IC and TSV technologies
  • Advanced Welding Techniques Analysis
  • Face and Expression Recognition
  • Corrosion Behavior and Inhibition
  • Copper Interconnects and Reliability
  • Nasolacrimal Duct Obstruction Treatments
  • Tensor decomposition and applications
  • Glaucoma and retinal disorders
  • Retinal Imaging and Analysis
  • Video Surveillance and Tracking Methods
  • Ocular Surface and Contact Lens
  • Medical Image Segmentation Techniques
  • Retinal Diseases and Treatments
  • Advanced Sensor and Energy Harvesting Materials
  • melanin and skin pigmentation
  • Advanced Clustering Algorithms Research
  • Concrete Corrosion and Durability
  • Remote-Sensing Image Classification
  • Non-Destructive Testing Techniques
  • Retinal and Macular Surgery
  • Surface Treatment and Residual Stress
  • Skin Protection and Aging

Chinese Academy of Medical Sciences & Peking Union Medical College
2022-2025

Sun Yat-sen University
2010-2024

South China University of Technology
2014-2024

Chinese Academy of Medical Sciences Dermatology Hospital
2024

Ministry of Agriculture and Rural Affairs
2023

Sun Yat-sen Memorial Hospital
2023

Chinese Academy of Fishery Sciences
2023

Foshan University
2020-2023

Baosteel (China)
2022

First Affiliated Hospital of Kunming Medical University
2021

Wire and arc additive manufacturing (WAAM) produces thin-walled parts superior to other methods, because of its high efficiency, good compactability, low cost. However, the WAAM accuracy is limited by large heat input. Here, 0.8 mm 316L stainless steel welding wire deposited via speed cold form 30-layered samples, with 2 thickness, up 65 height. The effects three process parameters (the bottom current mode, scanning speed, cooling time) on deposition stability, macro morphology, structure,...

10.3390/met9010109 article EN cc-by Metals 2019-01-21

10.1557/jmr.2003.0355 article EN Journal of materials research/Pratt's guide to venture capital sources 2003-11-01

Duplex stainless steel multi-pass welds were made at 0.15 MPa, 0.45 and 0.75 MPa pressure, simulating underwater dry hyperbaric welding by the flux-cored arc (FCAW) method, with of normal pressure as a benchmark. The purpose this work was to estimate effect ambient on microstructure, pitting corrosion resistance impact toughness weld metal. microstructure measurement revealed that ferrite content in metal is lowest, followed MPa. analysis potentiodynamic polarization tests 30 °C 50...

10.3390/ma10121443 article EN Materials 2017-12-18

This study aimed to predict the treatment outcomes in patients with diabetic macular edema (DME) after 3 monthly anti-vascular endothelial growth factor (VEGF) injections using machine learning (ML) based on pretreatment optical coherence tomography (OCT) images and clinical variables.An ensemble ML system consisting of four deep (DL) models five classical (CML) was developed posttreatment central foveal thickness (CFT) best-corrected visual acuity (BCVA). A total 363 OCT 7,587 data records...

10.21037/atm-20-1431 article EN Annals of Translational Medicine 2021-01-01

Unsupervised graph embedding aims to extract highly discriminative node representations that facilitate the subsequent analysis. Converging evidence shows a multiview provides more comprehensive relationship between nodes than single-view capture intrinsic topology. However, little attention has been paid excavating of each from heterogeneous networks in an unsupervised manner. To end, we propose novel method, called deep infomax (MVDGI). The backbone our proposed model sought maximize...

10.1109/tcyb.2022.3163721 article EN IEEE Transactions on Cybernetics 2022-04-15

Tensor spectral clustering (TSC) is an emerging approach that explores multi-wise similarities to boost learning. However, two key challenges have yet be well addressed in the existing TSC methods: (1) The construction and storage of high-order affinity tensors encode are memory-intensive hampers their applicability, (2) they mostly employ a two-stage integrates multiple different orders learn consensus tensor embedding, thus often leading suboptimal result. To this end, paper proposes...

10.1109/tpami.2024.3361912 article EN cc-by IEEE Transactions on Pattern Analysis and Machine Intelligence 2024-02-05

In previous studies, the 308-nm light-emitting diode (LED) has been proven safe and effective for treating vitiligo. However, direct comparisons between LED excimer lamp (308-nm MEL) treatment of vitiligo are lacking.

10.1111/phpp.12972 article EN Photodermatology Photoimmunology & Photomedicine 2024-05-01

Aims To develop a deep learning (DL) model for automatic classification of macular hole (MH) aetiology (idiopathic or secondary), and multimodal fusion network (MDFN) reliable prediction MH status (closed open) at 1 month after vitrectomy internal limiting membrane peeling (VILMP). Methods In this multicentre retrospective cohort study, total 330 eyes with 1082 optical coherence tomography (OCT) images 3300 clinical data enrolled from four ophthalmic centres were used to train, validate...

10.1136/bjophthalmol-2021-318844 article EN cc-by-nc British Journal of Ophthalmology 2021-08-04

Duplex stainless steel welded metals were underwater local dry prepared on S32101 lean duplex trapezoidal groove plates with a self-made drain cover employing Supercore 2205P flux-cored filler wire. Different heat inputs employed to investigate the effects mechanical characteristics and microstructure of metals. The results demonstrated that as was applied, austenite concentrations in weld increased. It found concentration fraction Σ3-austenite twin-grain boundaries followed same trends....

10.3390/ma16062289 article EN Materials 2023-03-13
Coming Soon ...