A5100380071- Particle accelerators and beam dynamics
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Advanced SAR Imaging Techniques
- Particle Accelerators and Free-Electron Lasers
- Superconducting Materials and Applications
- ICT Impact and Policies
- Advanced Neural Network Applications
- Remote-Sensing Image Classification
- Underwater Acoustics Research
- Ocean Waves and Remote Sensing
- Image Retrieval and Classification Techniques
- Geochemistry and Geologic Mapping
- Remote Sensing and Land Use
- Advanced Algorithms and Applications
- Brain Tumor Detection and Classification
- Infrared Target Detection Methodologies
- Muon and positron interactions and applications
- Cell Image Analysis Techniques
- Optical Systems and Laser Technology
- Vehicle License Plate Recognition
- Advanced Image Fusion Techniques
- Energy Harvesting in Wireless Networks
- Industrial Vision Systems and Defect Detection
- Domain Adaptation and Few-Shot Learning
- Metaheuristic Optimization Algorithms Research
Southwest Jiaotong University
2022-2024
Southeast University
2017-2024
Peking University
2022
Beijing Research Institute of Uranium Geology
2013-2021
China University of Geosciences (Beijing)
2017
Lanzhou Petrochemical Polytechnic
2011
Goethe University Frankfurt
2007
Synthetic Aperture Radar (SAR) has the advantage of continuous observation throughout day and in all weather conditions, is used a wide range military civil applications. Among these, detection ships at sea an important research topic. Ships SAR images are characterized by dense alignment, arbitrary orientation multiple scales. The existing algorithms unable to solve these problems effectively. To address issues, A YOLOV8-based oriented ship classification method using imaging with...
The detection of ships encompasses an abundance applications within the domains fishery management, marine rescue operations, and maritime monitoring. In recent years, a multitude detectors based on deep learning have been utilized for purpose ship using synthetic aperture radar (SAR) images. However, disturbed by strong scattering background land influence SAR target scale, existing face great challenges in detecting inshore small ships. To solve this problem, article proposes oriented...
With the rapid development of intelligent technologies, deep learning-based object detection has been widely applied in dynamic fields, especially vehicle management. However, challenges accurate recognition and tracking remain, particularly due to issues like intra-class variations inter-class similarities re-identification across camera viewpoints. This paper proposes a method for long-range small target based on YOLOv8 algorithm, aiming improve accuracy performance targets complex...
Model-based decomposition methods are widely used in full-polarization synthetic aperture radar (SAR), for the inversion and interpretation of ground features constitute an important approach understanding behavior backscattering. However, owing to substantial differences between land marine environments, different man-made natural vegetation scattering structures render existing models unable reasonably characterize scatterers on ships. Moreover, combination polarization neural networks...
Polarimetric synthetic aperture radar (PolSAR) as an active microwave imaging device employing pulse compression technology to obtain echo information of multiple polarization channels has been widely used in marine target detection. However, the detection ship targets PolSAR images is often disturbed by clutter interference, such side lobes, ghost ships, etc. Additionally, current common convolutional neural network object models are accompanied a large number parameters, thereby, their...
Fine-grained ship classification in optical remote sensing images is a major challenge the ocean observation field, elaborated as follows: First, cost of acquiring expensive. Obtaining numerous labeled samples difficult, resulting poor generalization ability training models. Second, features target cannot be accurately obtained owing to complex background interference. Third, inter-class similarity and intra-class diversity among different ships render difficult. In this study, we propose...
This paper serves as an overview to the special session "Neural Networks: Efficient Implementations and Applications" on IEEE International Conference ASIC (ASICON) 2017. Focusing state-of-the-art research progresses neural networks, our introduction consists of two main parts, namely efficient hardware implementations latest applications. Deep network (DNN) provides superior performance for complex tasks, which pushes one step further in many fields. However, development DNN, a lot...
As one of the main components injector II China ADS LINAC project, an RFQ working at 162.5MHz is used to accelerate proton beams 15mA from 30 keV 2.1 MeV. The four vane has been designed in collaboration with Lawrence Berkeley National Laboratory and built workshop Institute Modern Physics, Chinese Academy Sciences (IMP, CAS). Low power test cavity completed, it shows field flatness within ±1% unloaded Q 12600. RF conditioning results preliminary beam show output energy 2.16 MeV spread 3.5%...
Convolutional Neural Networks (CNNs) have been widely applied in various fields, such as image recognition, speech processing, well many big-data analysis tasks. However, their large size and intensive computation hinder deployment hardware, especially on the embedded systems with stringent latency, power, area requirements. To address this issue, low bit-width CNNs are proposed a highly competitive candidate. In paper, we propose an efficient, scalable accelerator for based parallel...
EUROTRANS (EUROpean Research Programme for the TRANSmutation of High Level Nuclear Waste in an Accelerator Driven System) is calling efficient high-current CW front-end accelerator system. A combination RFQ, normal conducting CH- (Crossbar H-mode) and super-conducting CH-DTL which aims to work at 352MHz accelerate a 30mA proton beam 17MeV has been studied as promising candidate. The preliminary conceptual study results are reported with respect dynamics design.
The MYRRHA (Multi-purpose hYbrid Research Reactor for High-tech Applications) Project is planned as an accelerator driven system (ADS) the transmutation of long-living radioactive waste. For this project a cw 4-Rod-RFQ with 176 MHz and total length about 4 m required. It supposed to accelerate protons from 30 keV up 1.5 MeV*. One main tasks during development RFQ very high reliability limit thermal stress inside reactor. Another challenge was compensate dipole component MYRRHA-RFQ which due...
Drill core is very important for researchers to study lithospheric structure. However, how save cores in good condition as well satisfy the need of studying drill anytime and anywhere we want are big questions. Thus, numerical images necessary. Hyperspectral have lots advantages, they can provide both true color hyperspectral messages. True be used direct observation study. And messages mineral identification. In practical use, there distortions scanned images, so correction method great...
The injector II RFQ accelerator of ADS is used to accelerate protons 10 mA from 35 keV 2.1 MeV. cavity structure the same as that SNS which has a square cross section, and it adopts π-mode rods enhance RF (radio frequency) stability cavity. Low power tests show flatness better than ±0.01 unloaded Q value 13000. CW (continuous wave) working condition was realized after long time conditioning Beam were conducted with current in pulse mode mode, respectively, indicates transmission efficiency...