A5000543331- Retinal Imaging and Analysis
- Spectroscopy and Chemometric Analyses
- Advanced Image Processing Techniques
- Advanced Glycation End Products research
- Spectroscopy Techniques in Biomedical and Chemical Research
- Advanced Image and Video Retrieval Techniques
- Gear and Bearing Dynamics Analysis
- Chaos-based Image/Signal Encryption
- Artificial Intelligence in Healthcare
- Speech and Audio Processing
- Automated Road and Building Extraction
- Advanced Image Fusion Techniques
- Natural Language Processing Techniques
- Inertial Sensor and Navigation
- Remote-Sensing Image Classification
- Image and Signal Denoising Methods
- Machine Learning and Data Classification
- Image Processing Techniques and Applications
- Structural Health Monitoring Techniques
- Spacecraft Design and Technology
- Optical measurement and interference techniques
- Image Enhancement Techniques
- Space Satellite Systems and Control
- Robotics and Sensor-Based Localization
- Advanced Text Analysis Techniques
Chengdu University of Technology
2025
Tan Kah Kee Innovation Laboratory
2024
Xiamen University
2024
National Yang Ming Chiao Tung University
2024
Beihang University
2024
Changchun University of Science and Technology
2023
Wuhan University
2013-2014
In this paper, we describe our systems in which the objective is to determine whether a given news article could be considered as hyperpartisan. Hyperpartisan that takes an extremely polarized political standpoint with intention of creating divide among public. We attempted several approaches, including n-grams, sentiment analysis, well sentence and document representation using pre-tained ELMo. Our best system pre-trained ELMo Bidirectional LSTM achieved accuracy 83% through 10-fold...
Model-based image denoising methods are well suited for use as processors in remote sensing systems such satellites due to their well-developed mathematical theory and low computational cost, but these can often only deal with a single type of random noise. In this paper, based on the model-based techniques, mixed noise algorithm using Gaussian curvature surface shearlet transform is proposed. The filtering (GCF) used suppress salt & pepper (SPN) get first denoised image. second obtained by...
Abstract In view of the intrinsic characteristics chaotic system, it can be widely applied in image encryption. Considering common issues insufficient performance and discontinuous range control parameters current systems, a hybrid model with simple easy to design is devised address these challenges. Through phase portraits, bifurcation diagrams, analysis Lyapunov exponent spectrum, demonstrated that proposed system exhibit robust traversal performance. To meet demands digital encryption, an...
Early and differential diagnosis of Alzheimer's disease (AD) has puzzled many clinicians. In this work, laser Raman spectroscopy (LRS) was developed to diagnose AD from platelet samples transgenic mice non-transgenic controls different ages. An adaptive Gaussian process (GP) classification algorithm used re-establish the models early AD, advanced control group with just two features capacity for noise reduction. Compared previous multilayer perceptron network method, GP showed much better...
Alzheimer's disease (AD) is a common form of dementia. Early and differential diagnosis AD has always been an arduous task for the medical expert due to unapparent early symptoms currently imperfect imaging examination methods. Therefore, obtaining reliable markers with clinical diagnostic value in easily assembled samples worthy significant. Our previous work laser Raman spectroscopy (LRS), which we detected platelet different ages transgenic mice non-transgenic controls, showed great...
The keyhole phenomenon is widely observed in laser materials processing, including welding, remelting, cladding, drilling, and additive manufacturing. Keyhole-induced defects, primarily pores, dramatically affect the performance of final products, impeding broad use these laser-based technologies. formation pores typically associated with dynamic behavior keyhole. So far, accurate characterization prediction features, particularly depth, as a function time has been challenging task. In situ...
Fundus photography (FP) is a crucial technique for diagnosing the progression of ocular and systemic diseases in clinical studies, with wide applications early screening diagnosis. However, due to nonuniform illumination imbalanced intensity caused by various reasons, quality fundus images often severely weakened, brings challenges automated screening, analysis, diagnosis diseases. To resolve this problem, we developed strongly constrained generative adversarial networks (SCGAN). The results...
Data debugging is to find a subset of the training data such that model obtained by retraining on has better accuracy. A bunch heuristic approaches are proposed, however, none them guaranteed solve this problem effectively. This leaves an open issue whether there exists efficient algorithm it To answer question and provide theoretical basis for further study developing algorithms debugging, we investigate computational complexity named Debuggable. Given machine learning $\mathcal{M}$ dataset...
Fundus photography (FP) is a crucial technique for diagnosing the progression of ocular and systemic diseases in clinical studies, with wide applications early screening diagnosis. However, due to non-uniform illumination imbalanced intensity caused by various reasons, quality fundus images often severely weakened, brings challenges automated screening, analysis diagnosis diseases. To resolve this problem, we developed strongly constrained generative adversarial networks (SCGAN). The results...
Image geo-localization estimates an image's global position by comparing it with a large-scale image database containing known positions. This localization technology can serve as alternative positioning method for unmanned aerial vehicles (UAV) in situations where system is unavailable. Feature-based image-matching methods typically involve descriptors constructed from pixel-level key points the images. The number of one be substantial. Filtering and these large quantities matching would...