A5063590605- Medical Image Segmentation Techniques
- Remote-Sensing Image Classification
- Image Retrieval and Classification Techniques
- Image and Signal Denoising Methods
- Face and Expression Recognition
- Advanced Image and Video Retrieval Techniques
- Radiomics and Machine Learning in Medical Imaging
- Remote Sensing and Land Use
- Advanced Neural Network Applications
- AI in cancer detection
- Metaheuristic Optimization Algorithms Research
- Brain Tumor Detection and Classification
- Advanced Image Fusion Techniques
- Image Processing Techniques and Applications
- Neural Networks and Applications
- Spectroscopy and Chemometric Analyses
- COVID-19 diagnosis using AI
- Lung Cancer Diagnosis and Treatment
- Advanced Clustering Algorithms Research
- Image Enhancement Techniques
- Image and Object Detection Techniques
- Advanced Image Processing Techniques
- IPv6, Mobility, Handover, Networks, Security
- Wireless Communication Networks Research
- Medical Imaging and Analysis
Kennesaw State University
2015-2024
Mercer University
2022
Taipei Medical University
2020-2021
Huazhong University of Science and Technology
2019
Anyang Normal University
2012-2017
Beihang University
2017
Southern Polytechnic State University
2005-2014
National Chung Hsing University
2012
University of Georgia
2012
Providence College
2008
Hyperspectral imaging fully portrays materials through numerous and contiguous spectral bands. It is a very useful technique in various fields, including astronomy, medicine, food safety, forensics, target detection. However, hyperspectral images include redundant measurements, most classification studies encountered the Hughes phenomenon. Finding small subset of effective features to model characteristics classes represented data for critical preprocessing step required render classifier...
Calculating and generating optimal motion path automatically is one of the key issues in virtual human planning. To solve point, improved A* algorithm has been analyzed realized this paper, we modified traditional by weighted processing evaluation function, which made searching steps reduced from 200 to 80 time 4.359s 2.823s feasible The artificial marker, can escape barrier trap effectively quickly, also introduced avoid invalid region repeatedly, making more effective accurate finding...
Early detection of lung cancer is an effective way to improve the survival rate patients. It a critical step have accurate nodules in computed tomography (CT) images for diagnosis cancer. However, due heterogeneity and complexity surrounding environment, it challenge develop robust nodule method. In this study, we propose two-stage convolutional neural networks (TSCNN) detection. The first stage based on improved U-Net segmentation network establish initial nodules. During stage, order...
Accurate volumetric segmentation of brain tumors and tissues is beneficial for quantitative analysis disease identification in multi-modal Magnetic Resonance (MR) images. Nevertheless, due to the complex relationship between modalities, 3D Fully Convolutional Networks (3D FCNs) using simple fusion strategies hardly learn nonlinear complementary information modalities. Meanwhile, indiscriminative feature aggregation low-level high-level features easily causes misalignment FCNs. On other hand,...
Fuzzy clustering model is an essential tool to find the proper cluster structure of given data sets in pattern and image classification. In this paper, a new weighted fuzzy C-Means (NW-FCM) algorithm proposed improve performance both FCM FWCM models for high-dimensional multiclass recognition problems. The methodology used NW-FCM concept mean from nonparametric feature extraction (NWFE) discriminant analysis (DAFE). These two concepts are combined unsupervised clustering. main features...
The integration of the transformer and convolutional neural network (CNN) has become a useful method for change detection in remote sensing images. main function is to capture global features while CNN more obtaining local features. However, such an not efficient very high-resolution (VHR) images with fine surface detail information. Hence, improve this traditional construction CNN, we propose dense Swin-Transformer-V2 (DST) VGG16, coined as DST-VGG, extracting discriminatory detection....
It is critical to have accurate detection of lung nodules in CT images for the early diagnosis cancer. In order achieve this, it necessary reduce false positive rate detection. Due heterogeneity and their similarity background, difficult distinguish true from numerous candidate nodules. this paper, solve challenging problem, we propose a Multi-Branch Ensemble Learning architecture based on three-dimensional (3D) convolutional neural networks (MBEL-3D-CNN). The method combines three key...
Speech plays an important role in human-computer emotional interaction. FaceNet used face recognition achieves great success due to its excellent feature extraction. In this study, we adopt the model and improve it for speech emotion recognition. To apply our work, signals are divided into segments at a given time interval, signal transformed discrete waveform diagram spectrogram. Subsequently, spectrogram separately fed end-to-end training. Our empirical study shows that pretraining is...
The goal of this research is to develop a process, using current imaging hardware and without human intervention, that provides an accurate timely detection alert concealed weapon its location in the image luggage. There are several processes existence able highlight or otherwise outline baggage but so far those still require highly trained operator observe resulting draw correct conclusions. We attempted three different approaches project. first approach uses edge combined with pattern...
Brain tissue segmentation in multi-modal magnetic resonance (MR) images is significant for the clinical diagnosis of brain diseases. Due to blurred boundaries, low contrast, and intricate anatomical relationships between regions, automatic without prior knowledge still challenging. This paper presents a novel 3D fully convolutional network (FCN) segmentation, called APRNet. In this network, we first propose anisotropic pyramidal reversible residual sequence (3DAPC-RRS) module integrate...
Unsupervised domain adaptation (UDA) methods have achieved promising performance in alleviating the shift between different imaging modalities. In this article, we propose a robust two-stage 3-D anatomy-guided self-training cross-modality segmentation (ASTCMSeg) framework based on UDA for unpaired image segmentation, including translation and stages. stage, first leverage similarity distributions patches to capture latent anatomical relationships an relation consistency (ARC) preserving...