Yuxuan Sun

ORCID: 0000-0001-7939-1072
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Asphalt Pavement Performance Evaluation
  • Domain Adaptation and Few-Shot Learning
  • Topic Modeling
  • Natural Language Processing Techniques
  • AI in cancer detection
  • Age of Information Optimization
  • Radiomics and Machine Learning in Medical Imaging
  • Advancements in Battery Materials
  • Infrastructure Maintenance and Monitoring
  • Finite Group Theory Research
  • Speech and dialogue systems
  • Coding theory and cryptography
  • Advanced Neural Network Applications
  • COVID-19 epidemiological studies
  • Digital Imaging for Blood Diseases
  • Innovative concrete reinforcement materials
  • COVID-19 diagnosis using AI
  • Supercapacitor Materials and Fabrication
  • graph theory and CDMA systems
  • Gut microbiota and health
  • Smart Materials for Construction
  • IoT and Edge/Fog Computing
  • Advanced Battery Materials and Technologies
  • Multimodal Machine Learning Applications
  • Industrial Vision Systems and Defect Detection

Aalto University
2023-2025

China University of Geosciences (Beijing)
2020-2025

Zhejiang University of Science and Technology
2022-2024

Rajamangala University of Technology Krungthep
2024

University of Hong Kong
2022-2024

Huazhong University of Science and Technology
2023-2024

Westlake University
2023-2024

Wuhan National Laboratory for Optoelectronics
2023-2024

University Town of Shenzhen
2023-2024

Tsinghua University
2023-2024

While Multiple Instance Learning (MIL) has shown promising results in digital Pathology Whole Slide Image (WSI) analysis, such a paradigm still faces performance and generalization problems due to high computational costs limited supervision of Gigapixel WSIs. To deal with the computation problem, previous methods utilize frozen model pretrained from ImageNet obtain representations, however, it may lose key information owing large domain gap hinder ability without image-level training-time...

10.1109/cvpr52729.2023.00720 article EN 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023-06-01

Cation-disordered Rocksalt oxides (DRXs) are a promising new class of cathode materials for Li-ion batteries due to their natural abundance, low cost and great electrochemical performance. High entropy strategy in Mn-based DRXs appears be an effective improving the rate capability, but it suffers from challenges including capacity degradation. The present paper reports group high (HE DRX) based on Ni2+ -Nb5+ pair; structural chemical evolution upon cycling with increasing transition metal...

10.1002/anie.202311930 article EN Angewandte Chemie International Edition 2023-09-04

Asphalt pavements are prone to cracking in low-temperature environments, and microwave heating (MH) can heal the cracks effectively. This research mainly investigates different MH effects on self-healing properties of asphalt mixtures. With this objective, three-point splitting test is conducted generate cracks. A oven employed heat samples, a thermal camera measures surface temperature. Results indicate that power time show positive linear correlation with healing efficiency, HI samples...

10.3390/buildings14051313 article EN cc-by Buildings 2024-05-07

In recent years, the potential of waste engine oil (WEO) as a rejuvenator for aged asphalt binders has gained significant attention. Despite this interest, understanding WEO's regeneration mechanism within binders, particularly its diffusion behavior when mixed with both and virgin remains limited. This study adopts molecular dynamics approach to constructing models virgin, aged, rejuvenated different WEO contents (3%, 6%, 9%, 12%). Key properties such density, glass transition temperature,...

10.3390/ma17102212 article EN Materials 2024-05-08

Accurate image classification and retrieval are of importance for clinical diagnosis treatment decision-making. The recent contrastive language-image pre-training (CLIP) model has shown remarkable proficiency in understanding natural images. Drawing inspiration from CLIP, pathology-dedicated CLIP (PathCLIP) been developed, utilizing over 200,000 text pairs training. While the performance PathCLIP is impressive, its robustness under a wide range corruptions remains unknown. Therefore, we...

10.1007/s10278-024-01128-4 article EN Deleted Journal 2024-07-09

ABSTRACT Hot spring microbial mats represent intricate biofilms that establish self‐sustaining ecosystems, hosting diverse communities which facilitate a range of biochemical processes and contribute to the structural functional complexity these systems. While community structuring across mat depth has received substantial attention, mechanisms shaping horizontal spatial composition structure remain understudied. Here, we explored contributions species source, local environment interaction...

10.1111/mec.17727 article EN Molecular Ecology 2025-03-15

Porous asphalt mixtures play a pivotal role in enhancing pavement drainage capacity and traffic safety, where the performance of binder constitutes determining factor. This study introduces an innovative advancement through development high-viscoelastic modifier corresponding modified based on SBS-modified asphalt, coupled with optimized preparation protocols. The optimal composition dosage were systematically determined standardized tests including penetration, ductility, softening point,...

10.3390/app15073631 article EN cc-by Applied Sciences 2025-03-26

The super absorbent polymer (SAP) has been attracting extensive concerns due to its strong capacity in water absorption and retention. amorphous hydrogels formed by the post-absorbent SAP have potential of clogging micro-cracks asphalt materials refraining rainwater from infiltrating. This provides possibility applying pavements seal or fill cracks relieve distresses caused infiltration underlying layers. Before exploring cracking sealing mechanism SAPs pavements, a series experiments were...

10.3390/ma16031082 article EN Materials 2023-01-26

Exploiting pseudo labels (e.g., categories and bounding boxes) of unannotated objects produced by a teacher detector have underpinned much recent progress in semi-supervised object detection (SSOD). However, due to the limited generalization capacity caused scarce annotations, often deviate from ground truth, especially those with relatively low classification confidences, thus limiting performance SSOD. To mitigate this problem, we propose dual pseudo-label polishing framework for Instead...

10.1609/aaai.v37i3.25455 article EN Proceedings of the AAAI Conference on Artificial Intelligence 2023-06-26

10.1016/j.tre.2024.103810 article EN Transportation Research Part E Logistics and Transportation Review 2024-10-30

Many lactic acid bacteria (LAB) are considered probiotics and their postbiotic compounds present additional health benefits through modulating the gut microbiota metabolites. This study aimed to explore modulation effect of potential postbiotics produced by Lactobacillus parabuchneri MF2103 (LPP), which is a food-isolated strain belongs LAB. In vitro fecal fermentation LPP was performed with Escherichia-Shigella- (E group) Faecalibacterium-enterotype inoculum (F group), relative abundance...

10.1016/j.lwt.2023.114782 article EN cc-by-nc-nd LWT 2023-04-28

Hierarchical federated learning (HFL) enables distributed training of models across multiple devices with the help several edge servers and a cloud server in privacy-preserving manner. In this paper, we consider HFL highly mobile devices, mainly targeting at vehicular networks. Through convergence analysis, show that mobility influences speed by both fusing data shuffling models. While is usually considered as challenge from perspective communication, prove it increases edge-level...

10.48550/arxiv.2401.09656 preprint EN other-oa arXiv (Cornell University) 2024-01-01

Single-crystal Ni-rich ternary cathode materials (811), with their comprehensive advantage of high energy density and low manufacturing cost, have been considered the most promising for Li-ion batteries. Due to strong O-p/TM-d covalent hybridization, complete oxidation Ni ions during electrochemical cycling is often accompanied by inevitable oxygen loss, resulting in an irreversible surface phase transition, accumulated lattice stress, poor thermal stability. Considering fact that...

10.1021/acsaem.4c01160 article EN ACS Applied Energy Materials 2024-08-05
Coming Soon ...