A5100380877- Medical Image Segmentation Techniques
- Advanced Image Processing Techniques
- Scheduling and Timetabling Solutions
- Color perception and design
- Image and Signal Denoising Methods
- Caching and Content Delivery
- Nietzsche, Schopenhauer, and Hegel
- Manufacturing Process and Optimization
- Cloud Data Security Solutions
- Classical Philosophy and Thought
- Image Enhancement Techniques
- Interactive and Immersive Displays
- Wound Healing and Treatments
- Advanced Image and Video Retrieval Techniques
- Advanced Data Storage Technologies
- Augmented Reality Applications
- Bee Products Chemical Analysis
- Engineering Applied Research
- Vibration and Dynamic Analysis
- Image Retrieval and Classification Techniques
- Industrial Vision Systems and Defect Detection
- Multimodal Machine Learning Applications
- Advanced Sensor and Control Systems
- Soft Robotics and Applications
- Cell Image Analysis Techniques
RAND Corporation
2024
Korea Advanced Institute of Science and Technology
2012-2023
Korea Institute of Science and Technology
2023
Chung-Ang University
2007-2021
Korea Research Institute of Ships and Ocean Engineering
2005-2020
Hanwha Solutions (South Korea)
2020
Daegu Gyeongbuk Institute of Science and Technology
2020
UCSI University
2019
Chosun University
2019
Hallym University
2008-2018
In this study, an artificial intelligence (AI) intrusion detection system using a deep neural network (DNN) was investigated and tested with the KDD Cup 99 dataset in response to ever-evolving attacks. First, data were preprocessed through transformation normalization for input DNN model. The algorithm applied refined preprocessing create learning model, entire used verify it. Finally, accuracy, rate, false alarm rate calculated ascertain efficacy of which found generate good results detection.
Mammalian spermatozoa must undergo capacitation, before becoming competent for fertilization. Despite its importance, the fundamental molecular mechanisms of capacitation are poorly understood. Therefore, in this study, we applied a proteomic approach identifying capacitation-related proteins boar order to elucidate events more precisely. 2-DE gels were generated from samples before- and after-capacitation. To validate results, Western blotting immunocytochemistry performed with 2...
With the increase of aging society worldwide, hospitals, medical practitioners and health insurers are now increasingly seeking ways to reduce cost healthcare while maitaining its quality. One solution subject attention from gouvernement providers is U-Health Smart Home that aims provide non-intrusive non-invasive monitoring assistance elderly directly in their own home. At POSTECH, project focused on building a smart home along with an autonomic system monitor as well inhabitants...
In this study, highly concentrated hydrogen nanobubble water was utilized as the blending for cement mortar to improve its compressive and flexural strengths. Highly nanobubbles can be obtained through osmosis. This concentration maintained by sustaining osmotic time. The specimens were cured 28 days, in which increased. improved their strength 2.25–13.48% 6.41–11.22%, compared those afforded plain water. densified at high concentrations, caused a decrease diameter. increased probability of...
The systems, processes, and organisms optimized for the evolution of nature over several hundred million years have helped designers architects seeking improved innovative solutions.Inspired by shape function nature, some created great architectures finding applying new design sources from nature's elements.The beautiful mysterious landscape surrounding us not only gives a appearance four seasons, but also provides source ideas human.When looking at elements such as trees, flowers, animals,...
We applied genetic algorithm to nurse scheduling problem. For time complexity problem of algorithm, we suggested efficient operators using a cost bit matrix which each cell indicates any violation constraints. A with 1 that the corresponding assignment violates constraints and needs no further consideration. The experimental results showed method generated faster in better quality compared traditional algorithm.
Inspired by the great success of deep neural networks in various fields computer vision, studies for image deblurring have begun to become more active recent days. However, most previous approaches often fail accurately remove blur artifacts, e.g., ghosting effects at object boundaries and degradation local details, restored results. In this paper, we propose a spectral-spatial network (DSSN) resolving problem single deblurring. Specifically, proposed method is able efficiently recover scene...
As cement manufacturing is associated with CO2 emissions that contribute to environmental problems, usage must be reduced and alternative aggregates developed. Thus, we developed evaluated a mortar uses feldspar as fine aggregate, which does not require additional mineral admixtures. Feldspar conventional plastering sand-based specimens were prepared aggregate proportion of 75%, 80%, or 85%. The compressive strength test results confirmed the feldspar-based was 1.1–4.5 times higher than...
Level set models are suitable for processing topological changes in different regions of images while performing segmentation. Active contour require an empirical setting initial parameters, which is tedious the end-user. This study proposes incremental level model with automatic initialization contours based on local and global fitting energies that enable it to capture image containing intensity corruption or other light artifacts. The region-based area length terms use signed pressure...
Image segmentation is a tedious task that suffers from constraints, such as blurred or weak edges and intensity inhomogeneity. Active contour models, including edge-based region-based methods, are extensively used for image segmentation. Each of these methods has its pros cons affect image-segmentation accuracy CPU processing time. This study combines local global fitting energies uses statistical information to drag contours toward object boundaries, thus overcoming The bias field, the...
The DHL began test and research on delivery system by drone. Because of growing demand about comercialized personal-use drone business, we needed a real-time unmanned control that individuals can afford. In this paper, propose an autonomous flight for We used marker recognition technique because it doesn't require high-spec device which cannot be afforded ordinary people. proposed maintains distance between in flight. estimates the calculating area recognized image. Performance is validated...
This paper proposes an integration of two neutrosophic based multi-criteria decision making methods, namely the data analytical hierarchy process (NDAHP) and Technique Order Preference by Similarity to Ideal Solution (TOPSIS) with maximizing deviation method, both on single-valued set (SVNS) evaluate efficiency general insurance companies in Malaysia. The level is a subjective vague matter, as can be further branched into operational efficiency, investment underwriting risk management...
This paper presents results from the third Thermal Image Super-Resolution (TISR) challenge organized in Perception Beyond Visible Spectrum (PBVS) 2022 workshop. The uses same thermal image dataset as first two challenges, with 951 training images and 50 validation at each resolution. A set of 20 was kept aside for testing. evaluation tasks were to measure PSNR SSIM between SR ground truth (HR noisy downsampled by four), also semi-registered HR (acquired another camera). outperformed those...
Time-shift is a crucial function of interactive televisions as like DVR and Internet TV broadcasting services. Automatic important event detection allows users utilize time- shift conveniently. In this paper, we propose method to extract events in baseball videos. the proposed method, first detect play scenes audio separately from video tracks. For robust scene extraction, off-line learning model having local adaptation based on ongoing analyzed video. And implemented with SVM-based...
목적 본 연구는 솔방울(pine cone)을 화장품 소재로서의 활용가능성을 확인하기 위하여 열수(WE), 주정(EE), 초임계방법(SFE)으로 추출하여 비교 실험을 진행하였다. 방법 각 추출방법으로 얻은 추출물의 파이토케미컬 성분을 확인하고자GC/MS로 분석하였고, 총 폴리페놀 함량과 2,2-diphenyl-1-picrylhydrazyl (DPPH) 라디칼소거능으로 항산화 활성을 측정하였다. 또한 항균 및 항염증 효과를 분석하였으며, 추출물을 에멀젼 제형에 첨가하여 안정성을 확인하였다. ê²°ê³¼ 함량 측정한 결과 WE는 180.75±1.83 mg/g, EE는 252.69±0.34 SFE는 270.10±0.97 mg/g이 확인되었다. DPPH법을 이용한 효과 측정 시 SFE로 추출물이 저농도(0.125%)에서도 높은 능력을 나타내었다. 솔방울 SEF 추출물은 SA에서 12.1±1.04 mm, SE에서 11.8±0.45 mm의 clear zone이 확인이 되어 기타 WE와...