A5100377910- Domain Adaptation and Few-Shot Learning
- Topic Modeling
- Advanced Neural Network Applications
- Multimodal Machine Learning Applications
- Natural Language Processing Techniques
- Brain Tumor Detection and Classification
- Phase Change Materials Research
- Educational Technology and Pedagogy
- Gear and Bearing Dynamics Analysis
- Metallurgy and Material Forming
- Image Processing Techniques and Applications
- Advanced Battery Technologies Research
- Adversarial Robustness in Machine Learning
- Advanced Decision-Making Techniques
- CCD and CMOS Imaging Sensors
- Translation Studies and Practices
- Radiation Detection and Scintillator Technologies
- Literature, Language, and Rhetoric Studies
- COVID-19 diagnosis using AI
- Advanced Computational Techniques and Applications
- Iterative Learning Control Systems
- Viral Infections and Vectors
- Medical Image Segmentation Techniques
- Geophysical Methods and Applications
- Speech Recognition and Synthesis
Inner Mongolia University
2022-2024
University of Shanghai for Science and Technology
2023-2024
Jinhua Academy of Agricultural Sciences
2010
Phase change materials (PCMs) have been used statically, which has caused the use of these to face challenges. Encapsulating PCMs and combining them with base fluid can significantly solve problem using in BTM systems. In present study, based on computational dynamics, forced convection heat transfer nano-encapsulated phase (NEPCM) a system are simulated. The main aim research is reduce temperature at surface hot cylinder. this research, we simulated lithium battery thermal management...
Few-shot image classification, whose goal is to generalize unseen tasks with scarce labeled data, has developed rapidly over the years. However, in traditional few-shot learning methods CNNs, non-local features and long-rang dependencies of may be lost, this leads a poor generalization trained model. With advantage self-attention mechanism Transformer, researchers have tried use vision transformer improve recently. these are more complicated take up lot computing resources, there no baseline...
Previous works trained the Transformer and its variants end-to-end achieved remarkable translation performance when there are huge parallel sentences available. However, these models suffer from data scarcity problem in low-resource machine tasks. To deal with mismatch between big model capacity of small training set, this paper adds BERT supervision on latent representation encoder decoder designs a multi-step algorithm to boost such basis. The includes three stages: (1) training, (2) (3)...
Computed tomography (CT) images play a vital role in diagnosing rib fractures and determining the severity of chest trauma. However, quickly accurately identifying large number CT is an arduous task for radiologists. We propose U-net-based detection method designed to extract fracture features at pixel level find rapidly precisely. Two modules are applied segmentation network-a combined attention module (CAM) hybrid dense dilated convolution (HDDC). The same layer encoder decoder fused...
Automatic detection of pitting on Ball Screw Drive (BSD) is essential to ensure normal production activities. However, the scarcity defective samples and precisely labeled data poses a significant challenge. To address this, we propose an efficient double self-supervised model that operates at both image pixel levels, aiming construct high-performance trained with defect-free for detecting unknown defects in BSD images. By incorporating global local information extracting features multiple...
In prior studies, domain adversarial neural networks (DANNs) are used to align image-level features regardless of foreground and background. However, the conventional discriminator in DANNs may leads feature extractor disregard cross-domain rather than aligning them. This phenomenon negatively impact classifier performance. We propose a novel loss reweighting technique that mitigates optimization conflict between classifier. The classification is reweighted based on prediction uncertainty...