A5079641793- Advanced Neural Network Applications
- Advanced Vision and Imaging
- Dynamics and Control of Mechanical Systems
- Advanced Algorithms and Applications
- 3D Surveying and Cultural Heritage
- Adaptive Control of Nonlinear Systems
- Robotics and Sensor-Based Localization
- Natural Language Processing Techniques
- Hydraulic and Pneumatic Systems
- Topic Modeling
- Advanced Text Analysis Techniques
- Medical Imaging and Analysis
- Image Enhancement Techniques
- Vibration Control and Rheological Fluids
- Distributed Control Multi-Agent Systems
- Vibration and Dynamic Analysis
- Text Readability and Simplification
- Image and Signal Denoising Methods
- Power Line Inspection Robots
- Luminescence Properties of Advanced Materials
- Gas Sensing Nanomaterials and Sensors
- Vehicle License Plate Recognition
- Text and Document Classification Technologies
- Advanced Decision-Making Techniques
- Soil Geostatistics and Mapping
University of Jinan
2020-2024
China Southern Power Grid (China)
2024
University of Edinburgh
2024
Hunan Institute of Engineering
2024
Sun Yat-sen University
2024
North China Electric Power University
2023
National Yang Ming Chiao Tung University
2022
Tencent (China)
2019-2021
Tianjin University
2016-2020
Beihang University
2013-2018
Predicting the popularity of online article sheds light to many applications such as recommendation, advertising and information retrieval. However, there are several technical challenges be addressed for developing best predictive capability. (1) The fluctuates under impacts external factors, which unpredictable hard capture. (2) Content meta-data features, largely determining content popularity, usually multi-modal nontrivial model. (3) Besides, it also needs figure out how integrate...
Real-time forwarding prediction for predicting online contents' popularity is beneficial to various social applications enhancing interactive behaviors. Cascade graphs, formed by propagation, play a vital role in real-time prediction. Existing cascade graph modeling methods are inadequate embed graphs that have hub structures and deep paths, or they fail handle the short-term outbreak of amount. To this end, we propose novel method includes an effective approach embedding variation sensitive...
A novel control method of addressing coupling and disturbance influences for finite-time energy-saving robust active suspension systems (ASSs) is investigated. By elaborately constructing effect indicators, the pros cons on ASSs are discussed, then a effects-triggered designed via second-order sliding mode technique. Importantly, good/bad effects assessed through well-designed nonlinear function. means determining if sign disturbances conforms to expected motion or not, addition beneficial...
Adversarial feature alignment is widely used in domain adaptive object detection. Despite the effectiveness on CNN-based detectors, its applicability to transformer-based detectors less studied. In this paper, we present AQT (adversarial query transformers) integrate adversarial into detection transformers. The generator a transformer which yields sequence of tokens, and discriminator consists novel token stack cross-attention layers. layers take as tokens from key-value pairs. Through...
This article provides an interesting exploration of character-level convolutional neural network solving Chinese corpus text classification problem. We constructed a large-scale language dataset, and the result shows that works better on than its corresponding pinyin format dataset. is first time applied to
In recent years, reconstructing a sparse map from simultaneous localization and mapping ( SLAM ) system on conventional CPU has undergone remarkable progress. However, obtaining dense the often requires high-performance GPU to accelerate computation. This paper proposes approach which can remove outliers obtain clean 3D model using in real-time. The processes keyframes establishes data association by multi-threading technology. are removed changing detections of associated vertices between...
Abstract The problem of fixed‐time formation control for a class second‐order nonlinear multi‐agent systems is studied. For such systems, algorithm proposed to maintain the connections among agents while avoiding collisions. Furthermore, radial basis function neural network used in design precisely approximate terms model. Then, dynamic sliding mode method suppress chattering phenomenon that may arise due control. A sufficient condition system achieve obtained by using different methods, as...
In this paper, we break down the large-scale project into small units so that can extract parameters relates most to total cost more easily. However, amount of these are huge, causing much trouble for RBF neural fitting, need reduce number by using PCA. Then put newly generated as inputs networks. The results empirical analysis showed a sound verification proposed estimation method is worthy expanding.
The lack of detection accuracy has been the pain point unattended substation inspection at all times. One algorithm in terms improved YOLOv5 is proposed paper so as to enhance accuracy. A backbone with unique attention mechanism designed extract more accurate feature maps. increases sensitivity model channel features by accurately location information relations and long-range dependencies a long range are encoded together spatial direction well other one preserved, helping locate objects....
Abstract The stable operation of multiple equipment in the converter station can ensure safety DC transmission system. Once fault occurs, it will pose a threat to and stability power grid. It is necessary monitor running state all kinds equipment, study monitoring system data based on sensor information three-dimensional positioning algorithm. In terms hardware design, TCP IP communication protocols are used connect sensors, sensors arranged according vibration conduction standard,...