• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.5
• /
• pp.2327-2347
• /
• 2018

Human actions can be recognized from depth sequences. In the proposed algorithm, we initially construct depth, motion maps (DMM) by projecting each depth frame onto three orthogonal Cartesian planes and add the motion energy for each view. The body part of the action (BPoA) is calculated by using bounding box with an optimal window size based on maximum spatial and temporal changes for each DMM. Furthermore, feature vector is constructed by using BPoA for each human action view. In this paper, we employed an ensemble based learning approach called Rotation Forest to recognize different actions Experimental results show that proposed method has significantly outperforms the state-of-the-art methods on Microsoft Research (MSR) Action 3D and MSR DailyActivity3D dataset.

• Journal of information and communication convergence engineering
• /
• v.17 no.2
• /
• pp.97-104
• /
• 2019

Distributed mobility management (DMM) does not use a centralized device. Its mobility functions are distributed among routers; therefore, the mobility services are not limited to the performance and reliability of specific mobility management equipment. The DMM scheme has been studied as a partially distributed architecture, which distributes only a packet delivery domain in combination with the software defined network (SDN) technology that separates the packet delivery and control areas. Particularly, a separated control area is advantageous in introducing a new service, thereby optimizing the network by recognizing the entire network situation and taking an optimal decision. The SDN-based mobility management scheme is studied as a method to optimize the packet delivery path whenever a mobile node moves; however, it results in excessive signaling processing cost. To reduce the high signaling cost, we propose a hybrid distributed mobility management method and analyze its performance mathematically.

• International journal of advanced smart convergence
• /
• v.7 no.4
• /
• pp.174-184
• /
• 2018

The efficiency and safety of social-overhead capital (SOC) public infrastructures have become an eminent social concern. In this regard, a continuous structural health monitoring has been widely implemented to oversee the robustness of such public infrastructures for the safety of the public. This paper deals with the analysis of a distributed mobility management (DMM) support for wireless sensor network (WSN) based information transmission system. The partial DMM support separates the data and control plane infrastructures, wherein, the control plane is managed by a particular mobility management network entity, while the data plane is distributed by the mobility anchors. The system will be able to optimize the information transmission for a wireless structural health monitoring of SOC public infrastructures specifically designed for bridges, and thus, guarantees the safety of public commuters.

• 한국ITS학회:학술대회논문집
• /
• 2008.11a
• /
• pp.309-310
• /
• 2008

• The Korean Journal of Physiology and Pharmacology
• /
• v.25 no.6
• /
• pp.517-523
• /
• 2021

The present study aims to investigate the impact of hydrogen-rich water on the lactic acid level in metformin-treated diabetic rats under hypoxia. Thirty Sprague-Dawley rats were randomly divided into five groups, including normal diet group, and diabetes model (DM) group, DM + metformin treatment (DMM) group, DMM + hypoxia treatment (DMMH) group and DMMH + hydrogen-rich water (DMMHR) group. We found that the levels of lactic acid, pyruvate and lactate dehydrogenase were significantly lower in the blood of DMMHR group than DMMH group. Superoxide dismutase and glutathione levels in liver and heart were significantly higher in DMMH group after hydrogen-rich water treatment, while malondialdehyde and oxidized glutathione levels were decreased in DMMHR group when compared with DMMH group, which indicates that hydrogen-rich water could reduce oxidative stress. qPCR analysis demonstrated that that pro-apoptotic genes Bax/Caspase-3 were upregulated in DM group and metformin treatment suppressed their upregulation (DMM group). However, hypoxic condition reversed the effect of metformin on apoptotic gene expression, and hydrogen-rich water showed little effect on these genes under hypoxia. HE staining showed that hydrogen-rich water prevented myocardial fiber damages under hypoxia. In summary, we conclude that hydrogen-rich water could prevent lactate accumulation and reduce oxidant stress in diabetic rat model to prevent hypoxia-induced damages. It could be served as a potential agent for diabetes patients with metformin treatment to prevent lactic acidosis and reduce myocardial damages under hypoxic conditions.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.258-260
• /
• 2022

In order to nurture talents who will lead the digital convergence era of the 4th industrial revolution that creates new knowledge and industries, research is being conducted on teaching methods that can improve the understanding of non-majors' SW concept, computational thinking ability, and convergence with majors is becoming Non-majors face difficulties in understanding and understanding the SW development environment, relevance to their major, and ability to converge. We used software education that is relatively easy to access for non-majors, and applied a demonstration-oriented model (DMM) that can be applied to beginners in SW education to understand the components and logical flow of ideas related to applications and majors used in real life. A convergence SW Learning method that combines repetitive implementation through instructor's demonstration and learner's modeling and learning motivational factors was proposed. In the experiment applying the teaching and learning method proposed in this paper, meaningful results were shown in terms of learning motivation and academic achievement in SW education.

• The Journal of the Petrological Society of Korea
• /
• v.21 no.2
• /
• pp.235-247
• /
• 2012

Peridotite xenoliths hosted by alkali basalts from South Korea occur in Baengnyeong Island, Jeju Island, Boeun, Asan, Pyeongtaek and Ganseong areas. K-Ar whole-rock ages of the basaltic rocks range from 0.1 to 18.9 Ma. The peridotites are dominantly lherzolites and magnesian harzburgites, and the constituent minerals are Fo-rich olivine ($Fo_{88.4-92.0}$), En-rich orthopyroxene, Di-rich clinopyroxene, and Cr-rich spinel (Cr# = 7.8-53.6). Hydrous minerals, such as pargasite and phlogopite, or garnet have not been reported yet. The Korean peridotites are residues after variable degree of partial melting (up to 26%) and melt extraction from fertile MORB mantle. However, some samples (usually refractory harzburgites) exhibit metasomatic enrichment of the highly incompatible elements, such as LREE. Equilibration temperatures estimated using two-pyroxene geothermometry range from ca. 850 to $1050^{\circ}C$. Sr and Nd isotopic compositions in clinopyroxene separates from the Korean peridotites show trends between depleted MORB-like mantle (DMM) and bulk silicate earth (BSE), which can be explained by secondary metasomatic overprinting of a precursor time-integrated depleted mantle. The Korean peridotite clinopyroxenes define mixing trends between DMM and EM2 end members on Sr-Pb and Nd-Pb isotopic correlation diagrams, without any corresponding changes in the basement. This is contrary to what we observe in late Cenozoic intraplate volcanism in East Asia which shows two distinct mantle sources such as a DMM-EM1 array for NE China including Baengnyeong Island and a DMM-EM2 array for Southeast Asia including Jeju Island. This observation suggests the existence of large-scale two distinct mantle domains in the shallow asthenosphere beneath East Asia. The Re-Os model ages on Korean peridotites indicate that they have been isolated from convecting mantle between ca. 1.8 and 1.9 Ga.

• Korean Journal of Food Science and Technology
• /
• v.43 no.1
• /
• pp.30-38
• /
• 2011

Effects of depolymerization treatments of a carbohydrase mixture (CM), ultrasound, and irradiation on the physical properties of defatted mustard meal-based edible films (DMM films) were investigated. DMM hydrocolloids were added to CM (0.42% (w/w solution)), treated by ultrasound (500-700 W, 10-30 min) or ${\gamma}$-ray (40-100 kGy) to prepare film-forming solutions. Films were formed by drying. The CM treatment at 0.42% (w/w), pH 5.5, and 40-$50^{\circ}C$ with a 0.5 hr incubation time resulted in the highest colloidal stability in the film-forming solution. The depolymerization treatments did not dramatically change the water vapor permeability of the films. The solubility of the film decreased up to 53.1% by the CM treatment. The ultrasound treatment (700 W-30 min) decreased tensile strength and elongation. The ultrasound treatment (600 W-20 min) resulted in more compact and uniform structures of the films. Flavor profiles were differentiated by the power level and the time of the ultrasound treatment.

• Journal of the Korean Geosynthetics Society
• /
• v.20 no.4
• /
• pp.43-52
• /
• 2021

As port development in soft ground is actively promoted for international logistics and transportation, the Deep Mixing Method (DMM) is continuously applied to form an improved column body directly in the ground by mixing cement with soil to secure the stability of the structure. However, in the case of cement, there is a problem of emitting a lot of greenhouse gases during the production process, so the development and use of new alternative materials are socially required to achieve the national goal of carbon neutrality. Accordingly, in this study, CMD-SOIL, developed to induce a hardening reaction similar to cement by recycling recycled resources, was used as a ground improvement material for the DMM. In addition, it was attempted to determine the possibility of replacing cement by conducting on-site test construction and evaluating applicability. As a result of the study, the compressive strength of CMD-SOIL compared to the design reference strength was 1.46 to 2.64 times higher in the field mixing test and 1.2 to 5.03 times higher than in the confirmed boring. In addition, the ratio (λ) of the compressive strength in the field to the design reference strength was 0.63 to 1.14, which was similar to the previous research results. Therefore, in the case of CMD-SOIL, it is possible to express the compressive strength necessary to secure stability, and there is no difference in applicability compared to existing materials such as ordinary portland cement and blast furnace slag cement, so it was analyzed that it could be used as a ground improvement material for the DMM.

• Journal of Broadcast Engineering
• /
• v.23 no.5
• /
• pp.614-621
• /
• 2018

This paper proposes a method of restoring corrupted depth image captured by depth camera through unsupervised learning using generative adversarial network (GAN). The proposed method generates restored face depth images using 3D morphable model convolutional neural network (3DMM CNN) with large-scale CelebFaces Attribute (CelebA) and FaceWarehouse dataset for training deep convolutional generative adversarial network (DCGAN). The generator and discriminator equip with Wasserstein distance for loss function by utilizing minimax game. Then the DCGAN restore the loss of captured facial depth images by performing another learning procedure using trained generator and new loss function.