• Journal of Electrical Engineering and Technology
• /
• v.13 no.6
• /
• pp.2468-2478
• /
• 2018

We propose the Sparse Feature Convolutional Neural Network (SFCNN) to reduce the volume of convolutional neural networks (CNNs). Despite the superior classification performance of CNNs, their enormous network volume requires high computational cost and long processing time, making real-time applications such as online-training difficult. We propose an advanced network that reduces the volume of conventional CNNs by producing a region-based sparse feature map. To produce the sparse feature map, two complementary region-based value extraction methods, cluster max extraction and local value extraction, are proposed. Cluster max is selected as the main function based on experimental results. To evaluate SFCNN, we conduct an experiment with two conventional CNNs. The network trains 59 times faster and tests 81 times faster than the VGG network, with a 1.2% loss of accuracy in multi-class classification using the Caltech101 dataset. In vehicle classification using the GTI Vehicle Image Database, the network trains 88 times faster and tests 94 times faster than the conventional CNNs, with a 0.1% loss of accuracy.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.6
• /
• pp.1029-1040
• /
• 2011

In this paper we propose an intelligent navigation algorithm for real world problem which can build a map without localization. Proposed algorithm operates online and furthermore does not require many memories for applying real world problem. After applying proposed algorithm to toy and huge data set, it does not require to calculate a whole eigenspace and need less memory compared to existing algorithm. Thus we can obtain that proposed algorithm is suitable for real world mobile navigation algorithm.

• Journal of Information Technology Applications and Management
• /
• v.28 no.4
• /
• pp.59-78
• /
• 2021

Mobile Grocery Shopping (MGS) has become the New Normal as the COVID-19 pandemic has changed the way consumers shop. Drawing on the framework of Customer Journey Map (CJM), this study explores consumers' MGS by identifying specific stages of Customer Journey and comparing consumers' satisfaction between PC-based online and mobile shopping experiences at each stage throughout the journey. This study collected 562 responses from subjects who have mobile and PC-based grocery shopping experiences at the major domestic e-Commerce platforms. Independent t-test analysis showed that differences in satisfaction between mobile and online shopping experiences exist in 5 main stages and 16 sub-stages of CJM. The results of service and technological innovation mentioned in the actual industry report were seen as empirical results leading to continued use of MGS as well as customer satisfaction. The findings of this study contribute to the research stream on Customer Journey by adopting the structure of CJM and analyzing specific stages of the journey in the context of MGS. Managerial implications for mobile-based business practitioners are also discussed.

• Journal of Korea Game Society
• /
• v.15 no.1
• /
• pp.35-44
• /
• 2015

The purpose of this study is to analyze user-generated rules in online games and to define the way of players' identities are developed. This study focuses on the online game League of Legends, which originated from use map system. Based on rule theory, this study analyzes the characteristics of user-generated rules. As the result, this study proves that the main rule of online game is not the rules made by developers, but the emergent rules made by players. Through rule-detection and rule-selection process, user-generated rules change the game system. Therefore, the result of this study expects to show meaningful roles of players in game system.

• Journal of Digital Contents Society
• /
• v.12 no.4
• /
• pp.559-565
• /
• 2011

In this paper, I propose a distributed game server system which has flexibility of enormous game character's processing in MMORPG using block scheme. I proposed game servers in distributed system grouped with zone which charge of each map, and the game servers in this zone divide game map into several unit blocks and process distributing blocks to game servers according to its performance. I proposed distributed game server system using block scheme can add/drop game map to/from easily through block process method in game and communication servers. Therefore, my proposed system can enormous character processing and add user created game map to existed game map easily in the future as well. Also, my scheme increases the system utilization and upgrades system stability by dynamic distribution of game map according to the number of characters and performance of servers.

• Journal of Korea Multimedia Society
• /
• v.15 no.3
• /
• pp.388-397
• /
• 2012

Recently, one of the received attraction fields in web based service is 'Human Relationship Service' that is called SNS. This relationship map service is able to deliver information to user more easily and visually because it is intuitive data that is linked with offline real world. While past map service put physical real information in the map, present map service is evolving into new communicative platform that expresses social relationship beyond simple search platform that shows real world. In this paper, we propose smart social grid system using sketch map that is based on online map service structure. This system has features such as standardized interface provision for various SNS, use to governance hub tool in case of establishing a personal network through expanded social grid, a role of bridge to mashup software linked with other SNS, user environment construction that reproduces social grid data, and the faster service setup by improved search technology.

• Journal of Broadcast Engineering
• /
• v.23 no.5
• /
• pp.642-655
• /
• 2018

Occupancy sensors installed in buildings and households turn off the light if the space is vacant. Currently PIR(pyroelectric infra-red) motion sensors have been utilized. Recently, the researches using camera sensors have been carried out in order to overcome the demerit of PIR that cannot detect stationary people. The detection of moving and stationary people is a main functionality of the occupancy sensors. In this paper, we propose an on-line human occupancy tracking method using convolutional neural network (CNN) and self-organizing map. It is well known that a large number of training samples are needed to train the model offline. To solve this problem, we use an untrained model and update the model by collecting training samples online directly from the test sequences. Using videos capurted from an overhead camera, experiments have validated that the proposed method effectively tracks human.

• Journal of Korea Game Society
• /
• v.7 no.2
• /
• pp.101-112
• /
• 2007

Along with the development of mobile technology, mobile game has undergone a change from downloadable game to network game, where a large number of users connect to server and play real-time game with a mobile phone. In order to develop these mobile network games economically, a proper, suitable game engine is needed. This study proposes the result of design and implementation of RWMMO-GE(Realtime Wireless Massively Multiplayer Online RPG Game Engine) that is used to develop a mobile network game. The structure of RWMMO-GE, which is the research result of this study, consists of major components such as Network/Client Module, Object Module, Map Tool, Script Editor, and Character Editor. The characteristics of the multiplayer mobile game developed by this engine is that a large number of players can play real-time game in a single map, which implies a possibility of a new business model in this area. This research is a result of the RWMMO-GE supported by 2006 IT Excellent Technology Support Project(No:A1300-0601-0125) of IITA(Institute Information Technology Advancement).

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.21 no.4
• /
• pp.601-607
• /
• 2012

The magnetic abrasive polishing (MAP), for online monitoring with AE sensor attachment, was performed in this study. To predict the surface roughness after the magnetic abrasive polishing of NAK80, the signal data acquired from the AE sensor were analyzed. A dimensionless coefficient, which consisted of average of AErms and standard deviation of AE signal, was defined as a characteristic of the MAP and a prediction model was obtained using least square method. A neural network, which had multiple input parameters from AE signals and polishing conditions, was applied for predicting the surface roughness. As a result of this study, it was seen that there was very close correlation between the AE signal and the surface roughness in the MAP. And then on-line prediction of the surface roughness after the MAP of the NAK80 was possible by the developed prediction model.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.41 no.4
• /
• pp.138-145
• /
• 2018

In the field of optimization, many studies have been performed on various types of Vehicle Routing Problem (VRP) for a long time. A variety of models have been derived to extend the basic VRP model, to consider multiple truck terminal, multiple pickup and delivery, and time windows characteristics. A lot of research has been performed to find better solutions in a reasonable time for these models with heuristic approaches. In this paper, by considering realtime traffic characteristics in Map Navigation environment, we proposed a method to manage realistic optimal path allocation for the logistics trucks and cargoes, which are dispersed, in order to realize the realistic cargo mixing allowance and time constraint enforcement which were required as the most important points for an online logistics brokerage service company. Then we developed a prototype system that can support above functionality together with delivery status monitoring on Map Navigation environment. First, through Map Navigation system, we derived information such as navigation-based travel time required for logistics allocation scheduling based on multiple terminal multiple pickup and delivery models with time constraints. Especially, the travel time can be actually obtained by using the Map Navigation system by reflecting the road situation and traffic. Second, we made a mathematical model for optimal path allocation using the derived information, and solved it using an optimization solver. Third, we constructed the prototype system to provide the proposed method together with realtime logistics monitoring by arranging the allocation results in the Map Navigation environment.