A5057313834- Image Processing Techniques and Applications
- Time Series Analysis and Forecasting
- Stock Market Forecasting Methods
- Advanced Image Fusion Techniques
- Advanced Image Processing Techniques
- Image and Signal Denoising Methods
- Cloud Data Security Solutions
- Advanced Vision and Imaging
- Image Enhancement Techniques
- Medical Image Segmentation Techniques
- Traffic Prediction and Management Techniques
- Remote-Sensing Image Classification
- Advanced Image and Video Retrieval Techniques
- Cryptography and Data Security
- Energy Load and Power Forecasting
- Optical measurement and interference techniques
- Remote Sensing and Land Use
- Advanced Neural Network Applications
- Privacy-Preserving Technologies in Data
- Advanced X-ray Imaging Techniques
- Anomaly Detection Techniques and Applications
- COVID-19 diagnosis using AI
- Retinal Imaging and Analysis
- AI in cancer detection
- Adversarial Robustness in Machine Learning
Hohai University
2024
Shandong Institute of Business and Technology
2020-2024
Wuhan University
2024
National University of Defense Technology
2023
Southwest Petroleum University
2022
Shandong University
2014-2021
Wuhan University of Technology
2021
Nanjing University of Information Science and Technology
2018-2020
Shandong University of Science and Technology
2016
Cloud storage represents the trend of intensive, scale and specialization information technology, which has changed technical architecture implementation method electronic records management. Moreover, it will p... | Find, read cite all research you need on Tech Science Press
Many keypoint detection and description methods have been proposed for image matching or registration. While these demonstrate promising performance single-modality matching, they often struggle with multimodal data because the descriptors trained on tend to lack robustness against non-linear variations present in data. Extending such requires well-aligned learn modality-invariant descriptors. However, acquiring is costly impractical many real-world scenarios. To address this challenge, we...
Drones have become prevalent robotic platforms with diverse applications, showing significant potential in Embodied Artificial Intelligence (Embodied AI). Referring Expression Comprehension (REC) enables drones to locate objects based on natural language expressions, a crucial capability for AI. Despite advances REC ground-level scenes, aerial views introduce unique challenges including varying viewpoints, occlusions and scale variations. To address this gap, we RefDrone, benchmark drone...
Large Language Models (LLMs) are progressively being utilized as machine learning services and interface tools for various applications. However, the security implications of LLMs, particularly in relation to adversarial Trojan attacks, remain insufficiently examined. In this paper, we propose TrojLLM, an automatic black-box framework effectively generate universal stealthy triggers. When these triggers incorporated into input data, LLMs' outputs can be maliciously manipulated. Moreover,...
This paper proposes a novel method for image magnification by exploiting the property that intensity of an varies along direction gradient very quickly. It aims to maintain sharp edges and clear details. The proposed first calculates low-resolution fitting surface with quadratic polynomial precision. Then, bicubic interpolation is used obtain initial gradients high-resolution (HR) image. are readjusted find constrained HR image, according spatial correlations between within local window. To...
Part of important structural edges in the image is smoothed due to small gradients, while others are preserved with greater gradients. Therefore, authors propose a two‐stage smoothing method based on edge‐patch histogram equalisation and patch decomposition. The authors' purpose increase gradient reducing texture region. they divide into edge‐patches where concentrated or non‐edge‐patches details by segmentation. needs be equalised histograms for increasing edge pixels. All patches...
An improved deflectometry for wavefront measurement using a composite fringe is proposed to reduce the projection fringes and improve accuracy. The single contains four in different directions. It goes through tested objects then captured by CCD camera. Two high frequency orthogonal patterns two period can be obtained from fast Fourier transform. unwrapping of wrapped phase accomplished corresponding heterodyne method. reconstructed integration partial derivatives. Using only one fringe,...
With the continuous development of deep learning, long sequence time-series forecasting (LSTF) has attracted more and attention in power consumption prediction, traffic prediction stock prediction. In recent studies, various improved models Transformer are favored. While these have made breakthroughs reducing time space complexity Transformer, there still some problems, such as predictive model being slightly lower than that Transformer. And ignore importance special values series. To solve...
The emergence of high-resolution (HR) remote sensing imagery showcases the continual advancements in technology but also sets higher demands for related tasks field, including image change detection. Due to their outstanding performance extracting salient features, convolutional neural networks (CNNs) have played a significant role and become widely utilized many computer vision tasks. encoder–decoder structure has confirmed effectiveness integrating multilevel feature information, as it...