A5015814459- Domain Adaptation and Few-Shot Learning
- Advanced Neural Network Applications
- Entrepreneurship Studies and Influences
- Parkinson's Disease Mechanisms and Treatments
- Advanced Image and Video Retrieval Techniques
- Topic Modeling
- Advanced Graph Neural Networks
- Advanced Algorithms and Applications
- Advanced Vision and Imaging
- Optical measurement and interference techniques
- Image and Video Stabilization
- Industrial Vision Systems and Defect Detection
- Higher Education Governance and Development
- Laser Applications in Dentistry and Medicine
- Adversarial Robustness in Machine Learning
- Fault Detection and Control Systems
- Image Enhancement Techniques
- Voice and Speech Disorders
- Vehicle License Plate Recognition
- Neonatal and fetal brain pathology
- Sleep and related disorders
- Sleep and Work-Related Fatigue
- Botulinum Toxin and Related Neurological Disorders
- ECG Monitoring and Analysis
- Advanced Data Storage Technologies
Xuzhou Medical College
2024-2025
Shenzhen Metro (China)
2024
Zhejiang University
2019-2023
Alibaba Group (China)
2023
State Key Laboratory of Industrial Control Technology
2020-2022
Tsinghua University
2015
In China, a significant number of undergraduates are experiencing poor sleep quality. This study was designed to investigate the prevalence quality and identify associated factors among in Jiangsu Province, China.
Knowledge graphs (KGs) store rich facts about the real world. In this paper, we study KG alignment, which aims to find alignment between not only entities but also relations and classes in different KGs. Alignment at entity level can cross-fertilize schema level. We propose a new approach, called DAAKG, based on deep learning active learning. With learning, it learns embeddings of entities, classes, jointly aligns them semi-supervised manner. estimates how likely an entity, relation or class...
In this paper, a data-driven model combining multivariate nonlinear chirp mode decomposition (MNCMD) with Granger causality (MGC) is proposed to analyze root causes for multiple plant-wide oscillations in process control system. Firstly, an MNCMD-based detector developed capture the oscillations, where oscillating variables caused by different sources are automatically clustered into various groups. Then, MGC applied each group obtain of oscillations. Compared state-of-the-art detection...
SSD <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">[2]</sup> has a very good performance in the field of object detection. When we use to detect objects image, produces number different scale prior boxes on feature maps. Then, it can base CNN extract features for final Classification and Regression. Though simplified process comparing Faster RCNN, FLOPs calculation storage size model are still enormous ordinary embedded devices. Parameter...
As convolutional neural network contains many redundant parameters, a lot of methods have been developed to compress the for accelerating inference. Among these, pruning, which is kind widely used approaches, can effectively decrease memory capacity and reduce computation cost. Herein, we propose competitive pruning approach based on Soft Filter Pruning (SFP) by taking account scaling factors y Batch Normalization (BN) layers as criterion filter selection strategy. During soft procedure, in...
Abstract According to the operation and management requirements of subway stations, equipment managed by station, such as lighting, escalators, broadcasting, etc., must be operated with one-button switch according a certain process every day. The first step in station is perform routine self-testing on these devices. Traditional operations mainly rely manual handling operators, which greatly affects their work efficiency. To optimize operational processes, it necessary establish intelligent...
Abdominal electrocardiogram is an important means to obtain fetal health condition during high-risk pregnancy. In this paper, a novel method for extracting heart rate from multi-channel mother abdomen electrocardiograms proposed using fast multivariate empirical mode decomposition technique (FMEMD). Firstly, FMEMD decomposes the multichannel ECG signals into set of modes. Two significant channels are selected according standard deviation fifth layer. Then continuous wavelet transform (CWT)...
Video Stabilization is one of the most important tasks in video surveillance on boarded airship platforms.Facing new problems that platforms usually have strong low-frequency vibration caused by wind and wobble streams could not be effectively reduced those method designed for casual handhold devices, a stabilization proposed.In this method, camera path estimation carried out modified using GPS attitude data key frames, planning given under 3 generalized conditions airship's motion...
In this paper, we present the "joint pre-training and local re-training'' framework for learning applying multi-source knowledge graph (KG) embeddings. We are motivated by fact that different KGs contain complementary information to improve KG embeddings downstream tasks. pre-train a large teacher embedding model over linked distill train student task-specific KG. To enable transfer across KGs, use entity alignment build subgraph connecting pre-trained target The is re-trained three-level...
Deep neural networks are proved to be very effective solve problems on image classification, object detection and segmentation. However, in cases where only limited hardware is acquired, it may a problem deploy big models with excellent performance as they sometimes calculation consuming. To overcome the limits power, memory calculation, channel pruning proposed compress model wise soon become common approach have compressed. Generally, three-stage pipeline containing training, finetuning....