A5100453210- Vehicle License Plate Recognition
- Handwritten Text Recognition Techniques
- Advanced Neural Network Applications
- Remote-Sensing Image Classification
- Metallurgical Processes and Thermodynamics
- Algorithms and Data Compression
- Remote Sensing and Land Use
- Magnetic and transport properties of perovskites and related materials
- Multiferroics and related materials
- Smart Cities and Technologies
- Video Surveillance and Tracking Methods
- AI in cancer detection
- Face and Expression Recognition
- Metallurgical and Alloy Processes
- Advanced Image Fusion Techniques
- Ferroelectric and Piezoelectric Materials
- Smart Parking Systems Research
- Machine Learning and Data Classification
- Adsorption and biosorption for pollutant removal
- Image Retrieval and Classification Techniques
- Image and Object Detection Techniques
- Text and Document Classification Technologies
- Anomaly Detection Techniques and Applications
- Heusler alloys: electronic and magnetic properties
- Digital Rights Management and Security
Shenzhen University
2025
University of Science and Technology Beijing
2015-2024
South China University of Technology
2024
Beijing Information Science & Technology University
2024
Zhejiang University of Technology
2023
Rice Research Institute
2023
Guangdong Academy of Agricultural Sciences
2023
Craft Group (China)
2022
Southwest University
2021
University of Tasmania
2021
A significant number of individuals have been affected by pandemic diseases, such as COVID-19 and seasonal influenza. Nucleic acid testing is a common method for identifying infected patients. However, manual sampling methods require the involvement numerous healthcare professionals. To address this challenge, we propose novel transoral swab robot designed to autonomously perform nucleic using visual-tactile fusion approach. The comprises series-parallel hybrid flexible mechanism precise...
In medical image segmentation, it is often necessary to collect opinions from multiple experts make the final decision. This clinical routine helps mitigate individual bias. However, when data annotated by experts, standard deep learning models are not applicable. this paper, we propose a novel neural network framework called Multi-rater Prism (MrPrism) learn segmentation labels. Inspired iterative half-quadratic optimization, MrPrism combines task of assigning multi-rater confidences and...
Previous license plate recognition (LPR) methods have achieved impressive performance on single-type plates. However, multi-type is still challenging due to various character layouts and fonts. There are two main problems: one that models prone incorrectly perceive the location of characters diverse layouts, other different categories may similar glyphs fonts, causing misidentification. Therefore, solve above problems, we propose plug-and-play modules based an attention-based framework for...
In recent years, object detection has shown excellent results on a large number of annotated data, but when there is discrepancy between the data and real test performance trained model often degraded it directly transferred to dataset. Compared with natural images, remote sensing images have great differences in appearance quality. Traditional methods need re-label all image before interpretation, which will consume lot manpower time. Therefore, practical significance study Cross-Domain...
Due to the angle variations especially in unconstrained scenarios, face detection and alignment have become challenging tasks. In existing methods, are always conducted separately, which can greatly increase computation cost. Moreover, this separation will abandon inherent correlation underlying two paper, we propose a simple but effective architecture, named Angle-Sensitivity Cascaded Networks (ASCN), for jointly conducting rotation-invariance alignment. ASCN mainly consists of three...
Various surface defects in automated fiber placement (AFP) processes affect the forming quality of components. In addition, defect detection usually requires manual observation with naked eye, which leads to low production efficiency. Therefore, automatic solutions for recognition have high economic potential. this paper, we propose a multi-scale AFP algorithm, named spatial pyramid feature fusion YOLOv5 channel attention (SPFFY-CA). The (SPFFY) adopts dilated convolutions (SPDCs) fuse maps...
Dimensionality reduction techniques can remove redundant information from hyperspectral images (HSIs) and improve discriminability. However, due to the inherent nonlinear characteristics of HSI, there may be non-Euclidean structures in data its topological properties make it suboptimal recover low-dimensional manifolds by means a linear projection. As result, projection high-dimensional space discriminative is not always effective. To better explore intrinsic geometric structure we propose...
Abstract Recently, magnetization switching driven by spin–orbit torque (SOT) has been intensely studied. However, it is still a challenge to effectively control the spin Hall angle (SHA) and critical current density for SOT switching. With help of multiferroic BiFeO 3 (BFO) thin films, method adjust SHA proposed. The BFO‐based heterostructures with opposite spontaneous polarization fields show huge changes in both perpendicular magnetic anisotropy SOT‐induced variation effective SHAs...
In this work, we demonstrated the idea in a epitaxial heterostructure of La0.7Sr0.3MnO3/La0.7Te0.3MnO3 (LSMO/LTMO) which exhibited related transition from typical p–n junction (hole-electron) with remarkable diode effect as at low temperature to p–p like (hole- small polaron) rectification. Besides, dependence under positive and negative voltage bias exhibit inverse variation trends. These results are associated major carrier rising LTMO. It indicates that different dopping variety physical...
License plate detection is the first and essential step of license recognition system still challenging in real applications, such as on-road scenarios. In particular, small-sized oblique plates, mainly caused by distant mobile camera, are difficult to detect. this work, we propose a novel applicable method for degraded via vehicle-plate relation mining, which localizes coarse-to-fine scheme. First, estimate local region around using relationships between vehicle plate, can greatly reduce...
For fine-grained classification, it is extremely difficult and costly to acquire the annotated data. Hence, some studies propose use web data for classification. However, contains tremendous noisy labels, which can affect classification results. Although many previous discard via sample selection, they also valid The denotes hard or mislabeled samples that enhance robustness of model. To solve above problems, we a novel method irrelevant from while keeping Specifically, divide into clean...