A5100458050- Aquatic Ecosystems and Phytoplankton Dynamics
- Topic Modeling
- Hydrology and Watershed Management Studies
- Generative Adversarial Networks and Image Synthesis
- Glaucoma and retinal disorders
- Vascular Malformations Diagnosis and Treatment
- Marine and coastal ecosystems
- Soil and Water Nutrient Dynamics
- Flood Risk Assessment and Management
- Advanced Neural Network Applications
- Domain Adaptation and Few-Shot Learning
- Vascular Malformations and Hemangiomas
- Retinal Diseases and Treatments
- Natural Language Processing Techniques
- Remote Sensing in Agriculture
- Land Use and Ecosystem Services
- Constructed Wetlands for Wastewater Treatment
- Medical Research and Treatments
- Cancer Cells and Metastasis
- Spinal Fractures and Fixation Techniques
- Retinal Imaging and Analysis
- Cancer Research and Treatments
- Urban Heat Island Mitigation
- Anomaly Detection Techniques and Applications
- Diet and metabolism studies
Sun Yat-sen University
2022-2025
Capital Medical University
2022-2025
Hangzhou Normal University
2025
Xuzhou Medical College
2022-2025
Beijing Chao-Yang Hospital
2023-2025
Anhui University of Technology
2025
Xinjiang Agricultural University
2023-2024
Shanghai Jiao Tong University
2020-2024
Zhejiang Chinese Medical University
2022-2024
Sixth Affiliated Hospital of Sun Yat-sen University
2022-2024
We present MMDetection, an object detection toolbox that contains a rich set of and instance segmentation methods as well related components modules. The started from codebase MMDet team who won the track COCO Challenge 2018. It gradually evolves into unified platform covers many popular contemporary not only includes training inference codes, but also provides weights for more than 200 network models. believe this is by far most complete toolbox. In paper, we introduce various features...
Thanks to the large pre-trained vision-language models (VLMs) like CLIP [37], we can craft a zero-shot classifier by discrete prompt design, e.g., confidence score of an image being "[CLASS]" be obtained using VLM provided similarity between and sentence "a photo [CLASS]". Furthermore, prompting shows great potential for fast adaptation VLMs downstream tasks if fine-tune soft prompts with few samples. However, find common failure that improper fine-tuning or learning extremely few-shot...
As the core area in transformation of Kunming into an international center city, studying changes ecological environment quality and causes Dianchi Lake Basin is great significance for its future optimization landscape pattern. This study based on Google Earth Engine (GEE) platform to calculate Remote Sensing Ecological Index (RSEI) from 1990 2020. Then we used Mann-Kendall mutation detection obtain time points when significant RSEI occurred. Finally, Geodetector MGWR models were combined...
Abstract The relationships of Klotho levels with cognition and dementia are poorly understood. This study aimed to investigate the association between cognitive function determine causality using Mendelian randomization (MR). Based on data from National Health Nutrition Survey (NHANES) 2011–2014, this consisted 1875 older adults aged 60–79 years. Cognitive was assessed by digit symbol substitution test (DSST). We performed weighted multivariable-adjusted linear regression assess...
Abstract As a typical sub-deep reservoir in the upper reaches of Yangtze River southwest region, Zhangjiayan Reservoir is also an important source drinking water. Exploring role microorganisms material cycle water bodies great significance for preventing exacerbation eutrophication reservoir. In this study, samples from overlying five points were collected four times spring (April), summer (July), autumn (November), and winter (January) 2022–2023 using gas-tight sampler. Physicochemical...
To explore the status of kinesophobia in patients with osteoporotic vertebral compression fractures and analyze influencing factors different profiles. Cross-sectional survey study. A total 245 who underwent surgical treatment at our Department Orthopedics between January 2023 March 2024 were selected. general data questionnaire Tampa Scale for Kinesiophobia used investigation. Latent profile analysis was conducted to categorise fractures, while univariate logistic regression performed...
Purpose To examine the associations of commonly-used serum lipid measures (high-density lipoprotein cholesterol (HDL-C), low-density (LDL-C), total (TC) and triglycerides (TG)) with glaucoma. Methods This prospective cohort study included 400 229 participants from UK Biobank. Cox regression restricted cubic spline models polygenic risk scores were employed to investigate between lipids Results Over a mean follow-up 14.44 years, 6868 (1.72%) developed Multivariate revealed that higher levels...
This report reviews the impact of significant changes in industry since publication 2nd Global Cryptoasset Benchmarking Study 2018. It provides novel insights into state cryptoasset industry, having gathered data from 280 companies 59 countries and across four main market segments – exchanges, payments, custody mining.Highlights report- Full-time equivalent (FTE) employee growth slowed considerably following late-2017 frenzy. Respondents all segments, reported year-on-year 21 per cent 2019,...
The consumption of disposable materials is booming with the rapid development urbanization and industrialization, which may inevitably cause release toxic harmful substances during use them in daily life. This study was to estimate element levels such as Beryllium (Be), Vanadium (V), Zinc (Zn), Manganese (Mn), Cadmium (Cd), Chromium (Cr), Nickel (Ni), Cobalt (Co), Antimony (Sb), Barium (Ba), Lead (Pb), Iron (Fe), Copper (Cu), Selenium (Se) leachate subsequently assess health risk exposure...
During the growth season, jujube trees are susceptible to infestation by leaf mite, which reduces fruit quality and productivity. Traditional monitoring techniques for mites time-consuming, difficult, subjective, result in a time lag. In this study, method based on particle swarm optimization (PSO) algorithm extreme learning machine estimation of chlorophyll content (SPAD) under mite was proposed. Initially, image data SPAD values orchards four severities were collected analysis. Six...
Although 2D generative models have made great progress in face image generation and animation, they often suffer from undesirable artifacts such as 3D inconsistency when rendering images different camera viewpoints. This prevents them synthesizing video animations indistinguishable real ones. Recently, 3D-aware GANs extend for explicit disentanglement of pose by leveraging scene representations. These methods can well preserve the consistency generated across views, yet cannot achieve...
Background: Kidney-Yang deficiency syndrome (KDS) is a group of diseases related to hypothalamic-pituitary-adrenal (HPA) axis and sexual dysfunction. The folium Epimedium brevicornu Maxim. (FEB) includes raw prepared slices, named RFEB PFEB, respectively. PFEB traditionally believed be good for tonifying kidney-Yang improving However, there are few studies comparing the pharmacological effects their underlying mechanisms. In this study, we aimed compare safety on HPA function. Additionally,...
Open-world survival games pose significant challenges for AI algorithms due to their multi-tasking, deep exploration, and goal prioritization requirements. Despite reinforcement learning (RL) being popular solving games, its high sample complexity limits effectiveness in complex open-world like Crafter or Minecraft. We propose a novel approach, SPRING, read the game's original academic paper use knowledge learned reason play game through large language model (LLM). Prompted with LaTeX source...
Thanks to the large pre-trained vision-language models (VLMs) like CLIP, we can craft a zero-shot classifier by "prompt", e.g., confidence score of an image being "[CLASS]" be obtained using VLM provided similarity measure between and prompt sentence "a photo [CLASS]". Therefore, shows great potential for fast adaptation VLMs downstream tasks if fine-tune prompt-based measure. However, find common failure that improper fine-tuning may not only undermine prompt's inherent prediction...