A5101044069- Advanced Image and Video Retrieval Techniques
- Video Surveillance and Tracking Methods
- Myxozoan Parasites in Aquatic Species
- Flood Risk Assessment and Management
- Non-Invasive Vital Sign Monitoring
- Actinomycetales infections and treatment
- Emotion and Mood Recognition
- Face and Expression Recognition
- Soil Geostatistics and Mapping
- Domain Adaptation and Few-Shot Learning
- Lipoproteins and Cardiovascular Health
- AI in cancer detection
- Technology and Security Systems
- Advanced Neural Network Applications
- Advanced Battery Materials and Technologies
- Handwritten Text Recognition Techniques
- COVID-19 Clinical Research Studies
- Cutaneous Melanoma Detection and Management
- Infectious Diseases and Mycology
- Vehicle License Plate Recognition
- Tropical and Extratropical Cyclones Research
- Color perception and design
- Smart Agriculture and AI
- Hydrology and Watershed Management Studies
- Image and Video Quality Assessment
Kent State University
2025
University of Louisiana at Lafayette
2022
Ton Duc Thang University
2022
Vietnam National University Ho Chi Minh City
2021
Ho Chi Minh City Medicine and Pharmacy University
2021
Nha Trang University
2021
VinUniversity
2021
University of Oulu
2021
Muroran Institute of Technology
2020
Hanoi Architectural University
2020
Existing works on visual counting primarily focus one specific category at a time, such as people, animals, and cells. In this paper, we are interested in everything, that is to count objects from any given only few annotated instances category. To end, pose few-shot regression task. tackle task, present novel method takes query image together with exemplar the predicts density map for presence of all interest image. We also adaptation strategy adapt our network test using introduce dataset...
Thermally regenerative batteries (TRBs) is an emerging platform for extracting electrical energy from low-grade waste heat (T < 130 °C). TRBs using ammonia-copper redox couple can store waste-heat in a chemical form that be later discharged to upon demand. Previous thermally ammonia battery (TRAB) demonstrations suffered poor conversion efficiency when benchmarked against thermoelectric generators (TEGs). In this work, we report the highest power density date TRAB (280 W m–2 at 55 °C) with...
Language prior plays an important role in the way humans detect and recognize text wild. Current scene recognition methods do use lexicons to improve performance, but their naive approach of casting output into a dictionary word based purely on edit distance has many limitations. In this paper, we present novel incorporate both training inference stage system. We generate list possible outcomes find one that is most compatible with visual appearance text. The proposed method leads robust...
Background: The COVID-19 pandemic posed a significant healthcare crisis due to the emergence of acute respiratory infections, leading severe complications such as thrombosis. Meanwhile, D-dimer index serves dependable routine diagnostic marker for Therefore, it is hypothesized that could serve potential prognostic infection. Objective: To assess association between levels and severity among patients in Vietnamese central coastal region. Methods: A retrospective cross-sectional study was...
Smartwatches have rapidly evolved towards capabilities to accurately capture physiological signals. As an appealing application, stress detection attracts many studies due its potential benefits human health. It is propitious investigate the applicability of deep neural networks (DNN) enhance decision-making through However, manually engineering DNN proves a tedious task especially in complex nature this phenomenon. To end, we propose optimized network training scheme using architecture...
In the recent years, researches in implementing visual-based Automatic Facial Expression Analysis (FEA) and human emotions estimation have been a great interest had widespread applications. However, vision-based visual systems face challenges such as poor-quality images due to low light conditions, poker-faces, contrast of expressions emotions. Because advantages thermal - unsensitive ambient we focus on developing emotion estimating system using thermal-based cues. It overcomes limits...
This study investigates the change in extreme rainfall and river flooding for a large basin due to climate during summer monsoon using ensemble dataset (d4PDF) coupled with Integrated Flood Analysis System (IFAS). Frequent severe causes significant damage Japan. Therefore, we aim provide useful information mitigate flood damage. The area is Ishikari River (IRB) Hokkaido, We used d4PDF 5-km downscaled data as input IFAS model. results showed that, given increase rainfall, discharges from IRB...
Recently, remotely obtaining the photoplethysmogram (PPG) signal to estimate blood volume pulse or heart rate of a human has become research topic which gains increasing attention. In contrast longstanding contact methods (e.g., electrocardiogram (ECG)), remote PPG can tackle same task with superior convenience and fewer physical constraints. It been proved in studies that affects change color intensity on some body parts such as face wrist. Leveraging this, we propose new method uses Intel...
Lowering the low-density lipoprotein cholesterol (LDL-c) level is important for reducing cardiovascular (CV) events. However, little known about management of LDL-c in older patients with type 2 diabetes mellitus (T2DM). This study investigated prevalence and factors associated non-achievement goals T2DM a very high risk CV diseases.
Saliency detection simulates human's perception in locating crucial regions, enabling further processing for many practical applications. Even though saliency prediction conventional 2D images and videos have been well developed, on 360° contents is still challenging. For each pixel the equirectangular frame, there will be corresponding surrounding pixels according to their spherical coordinate. Therefore, convolution method may induce certain inaccuracy attempt simulate humans perceive...
Abstract Between 2010 and 2015, nocardiosis outbreaks caused by Nocardia seriolae affected many permit farms throughout Vietnam, causing mass fish mortalities. To understand the biology, origin, epidemiology of these outbreaks, 20 N. strains collected from in four provinces South-Central Coast along with two Taiwanese strains, were analysed using genetics genomics. Pulsed-field gel electrophoresis identified a single cluster amongst all Vietnamese that was distinct strains. Like PFGE...
Semantic segmentation of nighttime images has become an interesting research topic recently. In this work, we focus on semantic object recognition for driving scenes. The paper proposes a method to adapt the models trained daytime scenes through twilight time. process, Pyramid Scene Parsing Network (PSPNet) model is suggested provide advanced framework pixel prediction. goal reduce cost human annotation by transferring knowledge from typical illumination conditions. Our and tested Cityscape...
Abstract Nocardia seriolae has caused significant fish losses in Asia and the Americas recent decades, including Vietnam, which witnessed devastating economic social impacts due to this bacterial pathogen. Surveillance strategies are urgently needed mitigate N. dissemination Vietnamese aquaculture mariculture industries. Whole-genome sequencing (WGS) offers highest level of resolution discriminate closely related strains determine their putative origin transmission routes. However, WGS is...