A5075576909- Face recognition and analysis
- Cardiovascular Health and Disease Prevention
- Cerebrovascular and Carotid Artery Diseases
- AI in cancer detection
- Cardiac Imaging and Diagnostics
- Infection Control and Ventilation
- Emotion and Mood Recognition
- Video Surveillance and Tracking Methods
- Radiomics and Machine Learning in Medical Imaging
- Retinal and Optic Conditions
- Medical Image Segmentation Techniques
- Face and Expression Recognition
- Advanced Neural Network Applications
- Prostate Cancer Diagnosis and Treatment
- Advanced Malware Detection Techniques
- Cardiovascular Disease and Adiposity
- Teaching and Learning Programming
- Leaf Properties and Growth Measurement
- Online Learning and Analytics
- Human Pose and Action Recognition
- Geotechnical Engineering and Soil Mechanics
- E-Learning and Knowledge Management
- Greenhouse Technology and Climate Control
- COVID-19 diagnosis using AI
- Advanced MRI Techniques and Applications
Guangxi University
2020-2024
Research Institute of Petroleum Exploration and Development
2024
City University of Hong Kong
2020-2024
Coronavirus 2019 has made a significant impact on the world. One effective strategy to prevent infection for people is wear masks in public places. Certain service providers require clients use their services only if they properly masks. There are, however, few research studies automatic face mask detection. In this paper, we proposed RetinaFaceMask, first high-performance single stage detector. First, solve issue that existing did not distinguish between correct and incorrect wearing...
Coronavirus disease 2019 has seriously affected the world. One major protective measure for individuals is to wear masks in public areas. Several regions applied a compulsory mask-wearing rule areas prevent transmission of virus. Few research studies have examined automatic face mask detection based on image analysis. In this paper, we propose deep learning single-shot light-weight detector meet low computational requirements embedded systems, as well achieve high performance. To cope with...
Facial micro-expressions (MEs) refer to brief spontaneous facial movements that can reveal a person's genuine emotion. They are valuable in lie detection, criminal analysis, and other areas. While deep learning-based ME recognition (MER) methods achieved impressive success, these typically require pre-processing using conventional optical flow-based extract motions as inputs. To overcome this limitation, we proposed novel MER framework self-supervised learning motion for (SelfME). the best...
Segmentation of the carotid section encompassing common artery (CCA), bifurcation and internal (ICA) from three-dimensional ultrasound (3DUS) is required to measure vessel wall volume (VWV) localized vessel-wall-plus-plaque thickness (VWT), shown be sensitive treatment effect. We proposed an approach combine a centerline extraction network (CHG-Net) dual-stream centerline-guided (DSCG-Net) segment lumen-intima (LIB) media-adventitia boundaries (MAB) 3DUS images. Correct arterial location...
While three-dimensional (3D) late gadolinium-enhanced (LGE) magnetic resonance (MR) imaging provides good conspicuity of small myocardial lesions with short acquisition time, it poses a challenge for image analysis as large number axial images are required to be segmented. We developed fully automatic convolutional neural network (CNN) called cascaded triplanar autoencoder M-Net (CTAEM-Net) segment scar from 3D LGE MRI. Two sub-networks were the left ventricle (LV) myocardium and then within...
Segmentation of carotid vessel wall is required in volume (VWV) and local vessel-wall-plus-plaque thickness (VWT) quantification the artery. Manual segmentation time-consuming prone to interobserver variability. In this paper, we proposed a convolutional neural network (CNN) segment common artery (CCA) from 3D ultrasound images. The CNN involves three U-Nets that segmented (3DUS) images axial, lateral frontal orientations. maps generated by were consolidated novel average (SAN) paper....
Coronavirus 2019 has made a significant impact on the world. One effective strategy to prevent infection for people is wear masks in public places. Certain service providers require clients use their services only if they properly masks. There are, however, few research studies automatic face mask detection. In this paper, we proposed RetinaFaceMask, first high-performance single stage detector. First, solve issue that existing did not distinguish between correct and incorrect wearing...
Objective: Vessel-wall-volume (VWV) and localized vessel-wall-thickness (VWT) measured from 3D ultrasound (US) carotid images are sensitive to anti-atherosclerotic effects of medical/dietary treatments. VWV VWT measurements require the lumen-intima (LIB) media-adventitia boundaries (MAB) at common internal arteries (CCA ICA). However, most existing segmentation techniques were capable automating only CCA segmentation. An approach segmenting MAB LIB ICA was required accelerate quantification....
Vessel-wall volume and localized three-dimensional ultrasound (3DUS) metrics are sensitive to the change of carotid atherosclerosis in response medical/dietary interventions. Manual segmentation media-adventitia boundary (MAB) lumen-intima (LIB) required obtain these is time-consuming prone observer variability. Although supervised deep-learning models have been proposed, training requires a sizeable manually segmented set, making larger clinical studies prohibitive.
As the critical dynamic parameters for soil, an extensive examination of elastic modulus Ed and damping ratio λ in coarse-grained soil is significant theoretical practical importance. Currently, there a scarcity experimental equipment methods measuring soils. Moreover, studies examining influence relative density on these soils are largely absent. To investigate behavior under varying densities, number triaxial tests were conducted specific using DYNTTS type test apparatus. The findings...
Targeted prostate biopsy guided by multiparametric magnetic resonance imaging (mpMRI) detects more clinically significant lesions than conventional systemic biopsy. Lesion segmentation is required for planning MRI-targeted biopsies. The requirement integrating image features available in T2-weighted and diffusion-weighted images poses a challenge lesion from mpMRI.A flexible efficient multistream fusion encoder proposed this work to facilitate the multiscale of multiple streams. A...
Images of green infrastructure (gardens, corridor, roofs and grasslands) large area can be captured processed to provide spatial temporal variation in colours plant leaves. This may indicate average growth over urban landscape (community gardens, corridor etc). Towards this direction, short technical note explores development a simple automated machine learning program that accurately segregate colors from In newly developed program, algorithm has been modified adapted give the proportion...
Automated segmentation of carotid lumen-intima boundary (LIB) and media-adventitia (MAB) by deep convolutional neural networks (CNN) from three-dimensional ultrasound (3DUS) images has made assessment monitoring atherosclerosis more efficient than manual segmentation. However, training CNN still requires LIB MAB. Therefore, there is a need to improve the efficiency develop strategies accuracy for serial atherosclerosis. One strategy reduce time increase interslice distance (ISD) between...
Small faults are developed in Nanpu 1-5 areas of Oilfield. Due to the influence seismic data resolution, it is difficult predict spatial distribution thin reservoirs. In addition, structure reservoir sand bodies unclear. this paper, firstly, based on seismic, logging, and production dynamic from work area, theoretical analysis conducted sequence stratigraphy, sedimentary petrology, sedimentology, development geology. On basis, an accurate geological model established geostatistical inversion...
Segmentation of carotid vessel wall is required in volume (VWV) and local vessel-wall-plus-plaque thickness (VWT) quantification the artery. Manual segmentation time-consuming prone to interobserver variability. In this paper, we proposed a convolution neural network segment common artery (CCA) from 3D ultrasound images. The CNN involves three U-Nets that segmented (3DUS) images axial, lateral frontal orientations. maps generated by were consolidated novel average (SAN) paper. experimental...
Internships help students connect what they have learned in the classroom to real world, and with access internships are more likely graduate secure employment. However, many unable find an internship by time graduate. This experience report describes a program where volunteer software engineers mentor as work on open-source projects summer, offered alternative traditional experience. We catalog considerations involved providing similar internship, describe our program's design, provide two...