• The Bulletin of The Korean Astronomical Society
• /
• v.42 no.1
• /
• pp.51.3-51.3
• /
• 2017

To investigate the connection between radio activity and AGN outflows, we present a study of ionized gas kinematics by using [O III] ${\lambda}5007$ emission line along the radio jet for six radio AGNs. These AGNs are selected based on the radioactivity (L1.4GHz ${\geq}$ 1039.8 erg s-1) as well as optical properties as type 2 AGNs. By using the high spatial resolution of the Red Channel Cross Dispersed Echellette Spectrograph at the Multiple Mirror Telescope, we investigate in detail the [O III] and stellar kinematics. We spatially resolve and probe the central AGN-photoionization sizes, which is important in understanding the structures and evolutions of galaxies. We find that the typical central AGN-photoionization sizes of our targets are in range of 1.8-3.8 kpc. We study the [O III] kinematics along the radio jets to test whether there is a link between gas outflows in the narrow-line region and radio jet emissions. Contrary to our expectation, we find no evidence that the gas outflows are directly connected to radio jet emission.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.14 no.2
• /
• pp.117-128
• /
• 1990

The purpose of this study is to find fitness of brassiere by means of body measuring value, its variation volume, the evaluation of perceptive image, and the evaluation of the sense of wearing. The results were as follows; 1) According to comparison for body measuring value before and after wearing brassiere, bust point (B.P.) height, lower bust height, upper chest circumference, chest circumference, and bust depth are increased, and lower chest circumference, bust point breadth, shoulder middle point - B.P., B.P. -under bust, and cup horizontal girth are decreased. 2) The difference of variation volume by material is not accepted. The part above $20\%$ at variation rate is under the region of the armpit, that is, the region connected arm from the back. 3) The subjects replied that they wore the brassiere in order to compensate the breast and needed to wear it regardless of thiness and obesity. They wore the brassiere in order to dress themselves in good shape, and felt that it put pressure upon the body, while it had nothing to do with adjusting bodily temperature and gave the sense of security. 4) The estimate of the sense of wearing by material is recognized as the difference of the attention at attentive level $1\%$. The multiple factor analysis of each item in the sense of wearing showed that the items which are explained over $90\%$ by common factors are '1. Unpreasant in touch', '2. The part of edge is haggard', '15. Not to be fit'.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.3
• /
• pp.61-69
• /
• 2010

This paper presents an efficient object selection algorithm by analyzing and detecting of human activity. Generally, when people point any something, they will put a face on the target direction. Therefore, the direction of the face and fingers and was ordered to be connected to a straight line. At first, in order to detect the moving objects from the input frames, we extract the interesting objects in real time using background subtraction. And the judgment of movement is determined by Principal Component Analysis and a designated time period. When user is motionless, we estimate the user's indication by estimation in relation to vector from the head to the hand. Through experiments using the multiple views, we confirm that the proposed algorithm can estimate the movement and indication of user more efficiently.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.6 no.9
• /
• pp.2118-2135
• /
• 2012

With the rapid development of heterogeneous emerging wireless technologies and numerous types of mobile devices, the need to support robust mobile video streaming based on the seamless handover in Future Internet is growing. To support the seamless handover, several IP-based mobility management protocols such as Mobile IPv6 (MIPv6), fast handover for the MIPv6 (FMIPv6), Hierarchical MIPv6 (HMIPv6) and Proxy Mobile IPv6 (PMIPv6) were developed. However, MIPv6 depreciates the Quality-of-Service (QoS) and FMIPv6 is not robust for the video services in heterogeneous emerging wireless networks when the Mobile Node (MN) may move to another visited network in contrast with its anticipation. In Future Internet, the possibility of mobile video service failure is more increased because mobile users consisting of multiple wireless network interfaces (WNICs) can frequently change the access networks according to their mobility in heterogeneous wireless access networks such as 3Generation (3G), Wireless Fidelity (Wi-Fi), Worldwide Interoperability for Microwave Access (WiMax) and Bluetooth co-existed. And in this environment, seamless mobility is coupled according to user preferences, enabling mobile users to be "Always Best Connected" (ABC) so that Quality of Experience is optimised and maintained. Even though HMIPv6 and PMIPv6 are proposed for the location management, handover latency enhancement, they still have limit of local mobility region. In this paper, we propose a robust mobile video streaming in Heterogeneous Emerging Wireless Systems. In the proposed scheme, the MN selects the best-according to an appropriate metric-wireless technology for a robust video streaming service among all wireless technologies by reducing the handover latency and initiation time when handover may fail. Through performance evaluation, we show that our scheme provides more robust mechanism than other schemes.

• Journal of Ocean Engineering and Technology
• /
• v.34 no.6
• /
• pp.394-405
• /
• 2020

To assess the station-keeping performance of floating structures in the Arctic region, the ice load should be considered along with other environmental loads induced by waves, wind, and currents. However, present methods for performance evaluation in the time domain are not effective in terms of time and cost. An ice load generation module is proposed based on the experimental data measured at the KRISO ice model basin. The developed module was applied to a time domain simulation. Using the results of a captive model test conducted in multiple directions, the statistical characteristics of ice loads were analyzed and processed so that an ice load corresponding to an arbitrary angle of the structure could be generated. The developed module is connected to commercial dynamic analysis software (OrcaFlex) as an external force input. Station-keeping simulation in the time domain was conducted for the same floating structure used in the model test. The mooring system was modeled and included to reflect the designed operation scenario. Simulation results show the effectiveness of the proposed ice generation module and its application to station-keeping performance evaluation. Considering the generated ice load, the designed structure can maintain a heading angle relative to ice up to 4°. Station-keeping performance is enhanced as the heading angle conforms to the drift direction. It is expected that the developed module will be used as a platform to verify station-keeping algorithms for Arctic floating structures with a dynamic positioning system.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.10
• /
• pp.1517-1524
• /
• 2022

In this paper, we consider a heterogeneous network (HetNet) consisting of one macro base station and multiple small base stations, and assume the coordinated multi-point transmission between the base stations. In addition, we assume that the channel between the base station and the user consists of path loss and Rayleigh fading. Under these assumptions, we present the energy efficiency (EE) achievable by the user for a given base station and we formulate an optimization problem of dynamic cell selection and transmit power allocation to maximize the total EE of the HetNet. In this paper, we propose an unsupervised deep learning method to solve the optimization problem. The proposed deep learning-based scheme can provide high EE while having low complexity compared to the conventional iterative convergence methods. Through the simulation, we show that the proposed dynamic cell selection scheme provides higher EE performance than the maximum signal-to-interference-plus-noise ratio scheme and the Lagrangian dual decomposition scheme, and the proposed transmit power allocation scheme provides the similar performance to the trust region interior point method which can achieve the maximum EE.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.7 no.3
• /
• pp.383-391
• /
• 2017

In image processing, image segmentation has been studied by various methods in a long time. Image segmentation is the process of partitioning a digital image into multiple objects and face detection is a typical image segmentation field being used in a variety of applications that identifies human faces in digital images. In this paper, we propose a method for extracting the contours of faces included in images. Using the Viola-Jones algorithm, to do this, we detect the approximate locations of faces from images. But, the Viola-Jones algorithm could detected the approximate location of face not the correct position. In order to extract a more accurate face region from image, we use skin color in this paper. In details, face region would be extracted using the analysis of horizontal and vertical histograms on the skin area. Finally, the face contour is extracted using snake algorithm for the extracted face area. In this paperr, a modified snake energy function is proposed for face contour extraction based snake algorithm proposed by Williams et al.[7]

• Atmosphere
• /
• v.30 no.1
• /
• pp.31-46
• /
• 2020

The aircraft observation campaign was performed to investigate thermodynamic conditions of snowfall cloud over the East Sea of Korean peninsula from 2 February to 16 March 2018. During this period, four snowfall events occurred in the Yeongdong region and three cases were analyzed using dropsonde data. Snowfall cases were associated with the passage of southern low-pressure (maritime warm air mass) and expansion of northern high-pressure (continental polar air mass). Case 1 and Case 2a were related to low-pressure systems, and Case 2b and Case 3 were connected with high-pressure systems, respectively. And their thermodynamic properties and horizontal distribution of snowfall cloud were differed according to the influence of the synoptic condition. In Case 1 and Case 2a, atmospheric layers between sea surface and 350 hPa contained moisture more than 15 mm of TPW with multiple inversion layers detected by dropsonde data, while the vertical atmosphere of Case 2b and Case 3 were dry as TPW 5 mm or less with a single inversion inversion layer around 750~850 hPa. However, the vertical distributions of equivalent potential temperature (θ_e) were similar as moist-adiabatically neutral condition regardless of the case. But, their values below 900 hPa were about 10 K higher in Case 1 and Case 2a (285~290 K) than in Case 2b and Case 3 (275~280 K). The difference in these values is related to the characteristics of the incoming air mass and the location of the snowfall cloud.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• Korean Journal of Heritage: History & Science
• /
• v.55 no.1
• /
• pp.63-84
• /
• 2022

One of the ancient nomadic cultures, the Saka is generally regarded as an important intermediary in the ancient Eurasian cultural network. This study is the reinterpretation of the excavations conducted on the Katartobe tombs site of the Saka culture through a joint three-year-long project by the National Research Institute of Cultural Heritage in Korea in collaboration with the Cultural Heritage Research Institute under the National Museum of the Republic of Kazakhstan. The main discussion of the study deals with the burial rituals performed by the community who built the Katartobe tombs by the comparison and review of the various researches on the Saka tombs based on the archaeological artifacts discovered during excavation. The research has shown that the Saka tribes maintained the tradition of burying domesticated animals, such as horses, with its owner and performed burial rituals which often involved the use of fire. The archaeological remains of the Saka also show that the burial rituals like these formed the key aspect of their cultural heritage. The archaeological discoveries also show that the Saka mourners built wooden cists under a single mound when they needed to bury multiple corpses at once and sustained the practice of excarnation when burying the bodies of those who died in the different periods of time. Some burials included a tomb passage which was used not only for carrying the deceased but also for a separate burial ritual. The main discussion of this study also deals with the remnants of bones of animals buried with their deceased owners in the same kurgan, as well as the animal species and their locations in the kurgan, resulting in the discovery of diverse meanings connected with them. The pottery buried in the tombs were largely ceremonial offering vessels, just like others excavated at nearby Saka tombs and located around the buried corpse's head facing toward the west. The excavation of the tombs also shows that two vessels were arranged at the corners of the coffin where the feet are located, revealing the characteristic features of the burial practices maintained by the tribe who built the Katartobe tombs. It may be too early to come to a definite conclusion on the burial practices of the Saka due to the relative lack of research on the kurgans across Central Asia. Excavations so far show that the kurgans clustered in a single archaeological site tend to display differences as well as uniformities. In conclusion, the ancient Central Asian tombs need more detailed surveys and researches to be able to make strides in an effort to restore the cultural heritage of the ancient Central Asian tribes who played a crucial role in the Eurasian cultural landscape.