• Journal of Advanced Navigation Technology
• /
• v.27 no.2
• /
• pp.190-196
• /
• 2023

Recently, following the development of LIDAR technology which can detect distance from the object, the interest for LIDAR based 3D object detection network is getting higher. Previous networks generate inaccurate localization results due to spatial information loss during voxelization and downsampling. In this study, we propose an attention-based convergence method and a camera-LIDAR convergence system to acquire high-level features and high positional accuracy. First, by introducing the attention method into the Voxel-RCNN structure, which is a grid-based 3D object detection network, the multi-scale sparse 3D convolution feature is effectively fused to improve the performance of 3D object detection. Additionally, we propose the late-fusion mechanism for fusing outcomes in 3D object detection network and 2D object detection network to delete false positive. Comparative experiments with existing algorithms are performed using the KITTI data set, which is widely used in the field of autonomous driving. The proposed method showed performance improvement in both 2D object detection on BEV and 3D object detection. In particular, the precision was improved by about 0.54% for the car moderate class compared to Voxel-RCNN.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.15 no.4
• /
• pp.26-41
• /
• 2012

It is necessary to manage, forecast and prepare agricultural production based on accurate and up-to-date information in order to cope with the climate change and its impacts such as global warming, floods and droughts. This study examined the applicability as well as challenges of the object-based image analysis method for developing a land cover image classification algorithm, which can support the fast thematic mapping of wide agricultural areas on a regional scale. In order to test the applicability of RapidEye's multi-temporal spectral information for differentiating agricultural land cover types, the integration of other GIS data was minimized. Under this circumstance, the land cover classification accuracy at the study area of Kimje ($1300km^2$) was 80.3%. The geometric resolution of RapidEye, 6.5m showed the possibility to derive the spatial features of agricultural land use generally cultivated on a small scale in Korea. The object-based image analysis method can realize the expert knowledge in various ways during the classification process, so that the application of spectral image information can be optimized. An additional advantage is that the already developed classification algorithm can be stored, edited with variables in detail with regard to analytical purpose, and may be applied to other images as well as other regions. However, the segmentation process, which is fundamental for the object-based image classification, often cannot be explained quantitatively. Therefore, it is necessary to draw the best results based on expert's empirical and scientific knowledge.

• Korean Journal of Cognitive Science
• /
• v.21 no.1
• /
• pp.157-175
• /
• 2010

Understanding dialogue participant's emotion is important as well as decoding the explicit message in human communication. It is well known that non-verbal elements are more suitable for conveying speaker's emotions than verbal elements. Written texts, however, contain a variety of linguistic units that express emotions. This study aims at analyzing components for constructing an emotion ontology, that provides us with numerous applications in Human Language Technology. A majority of the previous work in text-based emotion processing focused on the classification of emotions, the construction of a dictionary describing emotion, and the retrieval of those lexica in texts through keyword spotting and/or syntactic parsing techniques. The retrieved or computed emotions based on that process did not show good results in terms of accuracy. Thus, more sophisticate components analysis is proposed and the linguistic factors are introduced in this study. (1) 5 linguistic types of emotion expressions are differentiated in terms of target (verbal/non-verbal) and the method (expressive/descriptive/iconic). The correlations among them as well as their correlation with the non-verbal expressive type are also determined. This characteristic is expected to guarantees more adaptability to our ontology in multi-modal environments. (2) As emotion-related components, this study proposes 24 emotion types, the 5-scale intensity (-2~+2), and the 3-scale polarity (positive/negative/neutral) which can describe a variety of emotions in more detail and in standardized way. (3) We introduce verbal expression-related components, such as 'experiencer', 'description target', 'description method' and 'linguistic features', which can classify and tag appropriately verbal expressions of emotions. (4) Adopting the linguistic tag sets proposed by ISO and TEI and providing the mapping table between our classification of emotions and Plutchik's, our ontology can be easily employed for multilingual processing.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2014.02a
• /
• pp.221-221
• /
• 2014

The present work proposes an improved numerical simulator for design and modification of large area capacitively coupled plasma (CCP) processing chamber. CCP, as notoriously well-known, demands the tremendously huge computational cost for carrying out transient analyses in realistic multi-dimensional models, because electron dissociations take place in a much smaller time scale (${\Delta}t{\approx}10-8{\sim}10-10$) than time scale of those happened between neutrals (${\Delta}t{\approx}10-1{\sim}10-3$), due to the rf drive frequencies of external electric field. And also, for spatial discretization of electron flux (Je), exponential scheme such as Scharfetter-Gummel method needs to be used in order to alleviate the numerical stiffness and resolve exponential change of spatial distribution of electron temperature (Te) and electron number density (Ne) in the vicinity of electrodes. Due to such computational intractability, it is prohibited to simulate CCP deposition in a three-dimension within acceptable calculation runtimes (<24 h). Under the situation where process conditions require thickness non-uniformity below 5%, however, detailed flow features of reactive gases induced from three-dimensional geometric effects such as gas distribution through the perforated plates (showerhead) should be considered. Without considering plasma chemistry, we therefore simulated flow, temperature and species fields in three-dimensional geometry first, and then, based on that data, boundary conditions of two-dimensional plasma discharge model are set. In the particular case of SiH4-NH3-N2-He CCP discharge to produce deposition of SiNxHy thin film, a cylindrical showerhead electrode reactor was studied by numerical modeling of mass, momentum and energy transports for charged particles in an axi-symmetric geometry. By solving transport equations of electron and radicals simultaneously, we observed that the way how source gases are consumed in the non-isothermal flow field and such consequences on active species production were outlined as playing the leading parts in the processes. As an example of application of the model for the prediction of the deposited thickness uniformity in a 300 mm wafer plasma processing chamber, the results were compared with the experimentally measured deposition profiles along the radius of the wafer varying inter-electrode gap. The simulation results were in good agreement with experimental data.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.31 no.4
• /
• pp.173-181
• /
• 2018

Recently composite materials have dominated most engineering fields, owing to their better performance, increased durability and flexibility to be customized and designed for a specific required property. This has given them unprecedented superiority over conventional materials. With the help of the ever increasing computational capabilities of computers, researchers have been trying to develop accurate material models for the complex and integrated properties of these composites. This has led to advances in virtual testing of composite materials as a supplement or a possible replacement of laboratory experiments to predict the properties and responses of composite materials and structures. This paper presents a review on the complex multi-scale modelling framework of the virtual testing machines, which involve computational mechanics at various length-scales starting with nano-mechanics and ending in structure level computational mechanics, with a homogenization technique used to link the different length scales. In addition, the paper presents the features of some of the biggest integrated virtual testing machines developed for study of concrete, including a multiscale modeling scheme for the simulation of the constitutive properties of nanocomposites. Finally, the current challenges and future development potentials for virtual test machines are discussed.

• Journal of Intelligence and Information Systems
• /
• v.18 no.2
• /
• pp.157-175
• /
• 2012

It is difficult to evaluate the performance of process innovation of e-procurement which has large scale and complex processes. The existing evaluation methods for measuring the effects of process innovation have been mainly done with statistically quantitative methods by analyzing operational data or with qualitative methods by conducting surveys and interviews. However, these methods have some limitations to evaluate the effects because the performance evaluation of e-procurement process innovation should consider the interactions among participants who are active either directly or indirectly through the processes. This study considers the e-procurement process as a complex system and develops a simulation model based on MAS(Multi-Agent System) to evaluate the effects of e-procurement process innovation. Multi-agent based simulation allows observing interaction patterns of objects in virtual world through relationship among objects and their behavioral mechanism. Agent-based simulation is suitable especially for complex business problems. In this study, we used Netlogo Version 4.1.3 as a MAS simulation tool which was developed in Northwestern University. To do this, we developed a interaction model of agents in MAS environment. We defined process agents and task agents, and assigned their behavioral characteristics. The developed simulation model was applied to G2B system (KONEPS: Korea ON-line E-Procurement System) of Public Procurement Service (PPS) in Korea and used to evaluate the innovation effects of the G2B system. KONEPS is a successfully established e-procurement system started in the year 2002. KONEPS is a representative e-Procurement system which integrates characteristics of e-commerce into government for business procurement activities. KONEPS deserves the international recognition considering the annual transaction volume of 56 billion dollars, daily exchanges of electronic documents, users consisted of 121,000 suppliers and 37,000 public organizations, and the 4.5 billion dollars of cost saving. For the simulation, we analyzed the e-procurement of process of KONEPS into eight sub processes such as 'process 1: search products and acquisition of proposal', 'process 2 : review the methods of contracts and item features', 'process 3 : a notice of bid', 'process 4 : registration and confirmation of qualification', 'process 5 : bidding', 'process 6 : a screening test', 'process 7 : contracts', and 'process 8 : invoice and payment'. For the parameter settings of the agents behavior, we collected some data from the transactional database of PPS and some information by conducting a survey. The used data for the simulation are 'participants (government organizations, local government organizations and public institutions)', 'the number of bidding per year', 'the number of total contracts', 'the number of shopping mall transactions', 'the rate of contracts between bidding and shopping mall', 'the successful bidding ratio', and the estimated time for each process. The comparison was done for the difference of time consumption between 'before the innovation (As-was)' and 'after the innovation (As-is).' The results showed that there were productivity improvements in every eight sub processes. The decrease ratio of 'average number of task processing' was 92.7% and the decrease ratio of 'average time of task processing' was 95.4% in entire processes when we use G2B system comparing to the conventional method. Also, this study found that the process innovation effect will be enhanced if the task process related to the 'contract' can be improved. This study shows the usability and possibility of using MAS in process innovation evaluation and its modeling.

• Journal of the Korean Geographical Society
• /
• v.50 no.1
• /
• pp.73-90
• /
• 2015

This article focuses on geo-institutional differentiation and a multi-scalar analysis of emerging capitalist development in Laos. It discusses the impact of the Greater Mekong Subregion on new institutional economic and economic geographical arrangements. It demonstrates the usefulness of the varieties of Asian capitalism approach. The rubber industry was chosen to unravel emerging but various sub-national institutional arrangements linked to higher scale levels. Rubber is a growing agribusiness industry throughout the country, led by the insatiable demand from China. Overall, this study shows that the capitalist development of the rubber industry features much geo-institutional differentiation, due to the different strategies of Chinese, Thai and Vietnamese investors. Since Laos is still in transition from a state-led economy to something else, it is impossible at this to identify the exact number capitalisms. Yet, the evidence on rubber clearly lays bare the presence of multiple institutional arrangements. Without more inclusiveness, however, the implications for regional development are worrying. Exclusive arrangements will most likely lead to more uneven regional development and higher regional inequality. To refine theories on sub-national varieties of capitalism in developing countries it is instructive to consider more explicitly the notion of regional personal capitalisms and the complex interplay between national and regional states and relationships between capital accumulation and livelihood analyses.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.34 no.5
• /
• pp.535-544
• /
• 2016

Even though existing methods for orthophoto production using expensive aircraft are effective in large areas, they are drawbacks when dealing with renew quickly according to geographic features. But, as UAV(Unmanned Aerial Vehicle) technology has advanced rapidly, and also by loading sensors such as GPS and IMU, they are evaluates that these UAV and sensor technology can substitute expensive traditional aerial photogrammetry. Orthophoto production by using UAV has advantages that spatial information of small area can be updated quickly. But in the case of existing researches, images of same altitude are used in orthophoto generation, they are drawbacks about repetition of data and renewal of data. In this study, we targeted about small slope area, and by using low-end UAV, generated orthophoto and DEM(Digital Elevation Model) through different altitudinal images. The RMSE of the check points is σ_h = 0.023m on a horizontal plane and σ_v = 0.049m on a vertical plane. This maximum value and mean RMSE are in accordance with the working rule agreement for the aerial photogrammetry of the National Geographic Information Institute(NGII) on a 1/500 scale digital map. This paper suggests that generate orthophoto of high accuracy using a different altitude images. Reducing the repetition of data through images of different altitude and provide the informations about the spatial information quickly.

• Proceedings of the Korean Institute Of Construction Engineering and Management
• /
• 2007.11a
• /
• pp.791-796
• /
• 2007

As large-scale buildings, skyscrapers, and multi-purpose buildings recently increase in numbers dramatically, the internal space of such buildings becomes more and more large and complicated accordingly. Since such structures usually accommodate a number of random people, the potential possibilities of disastrous tragedies are high, and the rates of injury and physical damage caused by the complicated system of the building also increase as well. However, most of the shelter designs of the existing buildings are based on the specifications according to the assigned laws and involved regulations. In this case, only general criteria are referred to regardless of the characteristics of each structure while other disaster-related features are not taken into consideration sufficiently. Since any actual fire may cause a terrible calamity, in such plans, shelter inducement can be neither safe nor effective. Thus, this study examines and analyzes currently run disaster prevention systems and shelter inducement facilities with COEX Mall as its subject, and analyzes the responding system to each situation based on the fire scenarios by means of As-Is Model. Through this analysis, presented are the measures to solve the problems of current disaster prevention systems and to improve shelter inducement methods effectively.

• Journal of the Korean Society for Library and Information Science
• /
• v.57 no.1
• /
• pp.93-114
• /
• 2023

Information extraction can facilitate the intensive analysis of documents by providing semantic triples which consist of named entities and their relations recognized in the texts. However, most of the research so far has been carried out separately for named entity recognition and relation extraction as individual studies, and as a result, the effective performance evaluation of the entire information extraction systems was not performed properly. This paper introduces two models of end-to-end information extraction that can extract various entity names in clinical records and their relationships in the form of semantic triples, namely pipeline and joint models and compares their performances in depth. The pipeline model consists of an entity recognition sub-system based on bidirectional GRU-CRFs and a relation extraction module using multiple encoding scheme, whereas the joint model was implemented with a single bidirectional GRU-CRFs equipped with multi-head labeling method. In the experiments using i2b2/VA 2010, the performance of the pipeline model was 5.5% (F-measure) higher. In addition, through a comparative experiment with existing state-of-the-art systems using large-scale neural language models and manually constructed features, the objective performance level of the end-to-end models implemented in this paper could be identified properly.