A5001471247- Geotechnical Engineering and Underground Structures
- Random lasers and scattering media
- Geotechnical Engineering and Soil Stabilization
- Optical Coherence Tomography Applications
- Geotechnical Engineering and Analysis
- Technology and Security Systems
- Advanced Statistical Methods and Models
- Catalytic Processes in Materials Science
- Remote-Sensing Image Classification
- Solid-state spectroscopy and crystallography
- Advanced Steganography and Watermarking Techniques
- Neural Networks and Reservoir Computing
- Advanced Photocatalysis Techniques
- Coral and Marine Ecosystems Studies
- Digital Media Forensic Detection
- Nonlinear Optical Materials Research
- Topic Modeling
- Coastal wetland ecosystem dynamics
- Natural Language Processing Techniques
- Structural Load-Bearing Analysis
- Face and Expression Recognition
- Civil and Geotechnical Engineering Research
- Imbalanced Data Classification Techniques
- Advanced Graph Neural Networks
- Advanced Optical Sensing Technologies
Ministry of Transport
2024
Guizhou University
2024
Nanjing Xiaozhuang University
2024
China State Construction Engineering (China)
2023
Zhejiang Sci-Tech University
2021-2023
Beijing Jiaotong University
2021-2022
China University of Geosciences
2021
Mangroves play an important role in many aspects of ecosystem services. should be accurately extracted from remote sensing imagery to dynamically map and monitor the mangrove distribution area. However, popular extraction methods, such as object-oriented method, still have some defects for imagery, being low-intelligence, time-consuming, laborious. A pixel classification model inspired by deep learning technology was proposed solve these problems. Three modules were designed improve...
Ghost imaging is widely used in underwater active optical because of its simple structure, long distance, and non-local imaging. However, the complexity environment will greatly reduce quality ghost To solve this problem, an method based on generative adversarial networks proposed study. The generator network adopts U-Net with double skip connections attention module to improve reconstruction quality. In training process, total loss function sum weighted loss, perceptual pixel loss....
Down-sampling Fourier single-pixel imaging is typically achieved by truncating the spectrum, where exclusively low-frequency coefficients are extracted while discarding high-frequency components. However, truncation of spectrum can lead to an undesired ringing effect in reconstructed result. Moreover, original necessitated grayscale basis patterns for illumination. This requirement limits speed because digital micromirror devices (DMDs) generate at a lower refresh rate. In order solve above...
Investigations from past earthquakes have shown that underground subway stations can suffer excessive deformation under strong seismic loads, leading to the damage of critical components and collapse structures. This study presents results finite element analyses on installed different soil constraint conditions. The plastic hinge distribution characteristics cut cover double-storey three-storey are analyzed using method software ABAQUS. Combined with static analysis column sections, a...
Underwater ghost imaging based on deep learning can effectively reduce the influence of forward scattering and back water. With help data-driven methods, high-quality results be reconstructed. However, training underwater requires enormous paired datasets, which are difficult to obtain directly. Although Cycle-GAN method solves problem some extent, blurring degree fuzzy class datasets generated by is relatively unitary. To solve this problem, a few-shot image generative network proposed....
Large language models (LLMs) have garnered sub-stantial attention and significantly transformed the landscape of artificial intelligence, due to their human-like understanding generation capabilities. However, despite excellent capabilities, LLMs lack latest information are constrained by limited context memory, which limits effectiveness in many real-time applications that require up-to-date information, such as personal AI assistants. Inspired recent study on enhancing with infinite...