A5037089484- Numerical methods in engineering
- Crystallization and Solubility Studies
- X-ray Diffraction in Crystallography
- Advanced ceramic materials synthesis
- Geotechnical Engineering and Underground Structures
- Advanced Neural Network Applications
- Advanced biosensing and bioanalysis techniques
- Network Security and Intrusion Detection
- Chromosomal and Genetic Variations
- Anomaly Detection Techniques and Applications
- Advanced Nanomaterials in Catalysis
- Crystallography and molecular interactions
- Reproductive Biology and Fertility
- Aluminum Alloys Composites Properties
- Advancements in Battery Materials
- Advanced Malware Detection Techniques
- Photoreceptor and optogenetics research
- Neuroscience and Neural Engineering
- Forest, Soil, and Plant Ecology in China
- Genetic diversity and population structure
- Nanocluster Synthesis and Applications
- Geophysical Methods and Applications
- Ovarian cancer diagnosis and treatment
- Breast Cancer Treatment Studies
- Advanced Decision-Making Techniques
Tiangong University
2022-2025
Guangzhou Medical University
2022-2025
Third Affiliated Hospital of Southern Medical University
2020-2025
State Key Laboratory of Respiratory Disease
2022-2025
Sixth Affiliated Hospital of Sun Yat-sen University
2024-2025
Sun Yat-sen University
2006-2025
Nanjing Normal University
2025
Beijing National Laboratory for Molecular Sciences
2024
Chinese Academy of Sciences
2024
University of Chinese Academy of Sciences
2024
Real-world applications have high demands for semantic segmentation methods. Although has made remarkable leap-forwards with deep learning, the performance of real-time methods is not satisfactory. In this work, we propose PP-LiteSeg, a novel lightweight model task. Specifically, present Flexible and Lightweight Decoder (FLD) to reduce computation overhead previous decoder. To strengthen feature representations, Unified Attention Fusion Module (UAFM), which takes advantage spatial channel...
Being effective and efficient is essential to an object detector for practical use. To meet these two concerns, we comprehensively evaluate a collection of existing refinements improve the performance PP-YOLO while almost keep infer time unchanged. This paper will analyze empirically their impact on final model through incremental ablation study. Things tried that didn't work also be discussed. By combining multiple refinements, boost PP-YOLO's from 45.9% mAP 49.5% COCO2017 test-dev. Since...
We propose a lightweight CPU network based on the MKLDNN acceleration strategy, named PP-LCNet, which improves performance of models multiple tasks. This paper lists technologies can improve accuracy while latency is almost constant. With these improvements, PP-LCNet greatly surpass previous structure with same inference time for classification. As shown in Figure 1, it outperforms most state-of-the-art models. And downstream tasks computer vision, also performs very well, such as object...
The better accuracy and efficiency trade-off has been a challenging problem in object detection. In this work, we are dedicated to studying key optimizations neural network architecture choices for detection improve efficiency. We investigate the applicability of anchor-free strategy on lightweight models. enhance backbone structure design neck, which improves feature extraction ability network. label assignment loss function make training more stable efficient. Through these optimizations,...
Direct application of metallic lithium (Li) as the anode in rechargeable metal batteries (LMBs) is still hindered by some annoying issues such dendrites formation, low Coulombic efficiency, and safety concerns arising therefrom. Herein, an advanced composite separator prepared facilely blade coating lightweight thin functional layers on commercial 12 µm polyethylene to stabilize Li anode. The simultaneously improves ion transport deposition behaviors with uniform distribution properties,...
In this paper, a second-order backward differentiation formula compact difference scheme with the truncation error of order 1+α(0<α<1) for time and 4 space to fractional-order Volterra equation is considered. The integral term treated by means convolution quadrature suggested Lubich fourth-order accuracy approximation applied derivative. stability convergence in new norm are proved energy method. Numerical experiments that total agreement our analysis reported.
The loss of function AAA (ATPases associated with diverse cellular activities) mechanoenzymes has been linked to diseases, and small molecules that activate these proteins can be powerful tools probe mechanisms test therapeutic hypotheses. Unlike chemical inhibitors bind a single conformational state block enzyme function, activator binding must permissive different states needed for mechanochemistry. However, we do not know how activated by molecules. Here, focus on valosin-containing...
Due to the size effects of rockfill materials, settlement difference between numerical simulation and in situ monitoring dams is a topic general concern. The constitutive model parameters obtained from laboratory triaxial tests often underestimate deformation high dams. Therefore, by back analysis were used calculate predict long-term Instead using artificial neural networks (ANNs), response surface method (RSM) was employed replace finite element optimization iteration. Only 27 training...
<title>Abstract</title> <bold>Background.</bold> Robotic surgery for breast cancer is increasingly being adopted, but the evidence on its effectiveness and safety remains limited. This study aimed to compare surgical quality short-term outcomes of robotic endoscopic axillary lymph node dissection in patients with node-positive cancer. Here we report this trial. <bold>Methods.</bold> single-center retrospective compared efficacy (E-ALND) (R-ALND) Patients were recruited from those who...