• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.13 no.3
• /
• pp.66-77
• /
• 2014

UTIS(Urban Traffic Information System) directly collects link travel time in urban area by using probe vehicles. Therefore it can estimate more accurate link travel speed compared to other traffic detection systems. However, UTIS includes some missing data caused by the lack of probe vehicles and RSEs on road network, system failures, and other factors. In this study, we suggest a new model, based on k-NN algorithm, for imputing missing data to provide more accurate travel time information. New imputation model is an adaptive k-NN which can flexibly adjust the number of nearest neighbors(NN) depending on the distribution of candidate objects. The evaluation result indicates that the new model successfully imputed missing speed data and significantly reduced the imputation error as compared with other models(ARIMA and etc). We have a plan to use the new imputation model improving traffic information service by applying UTIS Central Traffic Information Center.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.798-801
• /
• 2014

Nowadays, the demand for multimedia services has grown with the rapid evolution in the digital era. But due to external causes in the process of processing, transmitting and storing image data, the images are damaged. One of the major causes of such damage is known to be noise. Some of the most commonly used methods for removing noise are CWMF(center weighted median filter), A-TMF(alpha-trimmed mean filter) and AWMF(adaptive weighted median filter). However, these filters all leave a bit to be desired in removing noise in a complex noise environment. Therefore this paper suggest an image restoration filter algorithm that first judges the noise and sets a adjustment weight based on the median value and distance of the mask to remove the complex noise. For an objective analysis, the results were compared against existing methods and the PSNR(peak signal to noise ratio) was used as a reference.

• Journal of Broadcast Engineering
• /
• v.7 no.1
• /
• pp.55-65
• /
• 2002

Wavelet transform has recently emerged as a promising technique for video processing applications due to its flexibility in representing non-stationary video signals. Motion estimation which uses wavelet transform of octave band division method is applied In many places but if motion estimation error happens in the lowest frequency band. motion estimation error is accumulated by next time strep and there has the Problem that time and the data amount that are cost In calculation at each steps are increased. On the other hand. wavelet packet that achieved the best image quality in a given bit rate from a rate-distortion sense is suggested. But, this method has the disadvantage of time costs on designing wavelet packet. In order to solve this problem we solved this problem by introducing Top_down method. But we did not find the optimum solution in a given butt rate. That image variance can represent image complexity is considered in this paper. In this paper. we propose a fast multiresolution motion estimation scheme based on the adaptive wavelet transform for video compression.

• Journal of the Korea Society for Simulation
• /
• v.15 no.4
• /
• pp.11-19
• /
• 2006

With the emergence of pervasive computing, personalization becomes an important issue to provide with users customized services, anywhere and anytime in their specific environment. Many researches have shown the possibilities of personalization by acquiring and processing sensor information around users. However, personalization remains still at its infancy, since most researches have failed to consider various contexts comprehensively besides sensor data, and just developed tailored services for a specific application domain. In this work, we propose a simulation-based CAS (context Aware Web Services) development methodology. Our methodology considers various contexts on users (eg. current location), web services (eg. response time), devices (eg. availability) and environment (eg. sensor data) all together by simulating them on the fly for personalized and adaptable services.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.8
• /
• pp.1851-1856
• /
• 2013

An XML is widely used for web services, such as SOAP(Simple Object Access Protocol) and REST (Representational State Transfer), and also de facto standard for representing data. Since the XML parser using DOM(Document Object Model) requires a preprocessing task creating a DOM-tree, and then storing it into memory, embedded systems with limited resources typically employ a streaming XML parser without preprocessing. In this paper, we propose a new architecture for the streaming XML parser using an APSearch(Adaptive Parallel Search) on FPGA(Field Programmable Gate Array). Compared to other approaches, the proposed APSearch parser dramatically reduces overhead on the software side and achieves about 2.55 and 2.96 times improvement in the time needed for an XML parsing. Therefore, our APSearch parser is suitable for systems to speed up XML parsing.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.12A
• /
• pp.1182-1188
• /
• 2006

Adaptive OFDM(Orthogonal Frequency Division Multiplexing) improves data capacity and system performance over multipath fading by adaptively changing modulation schemes according to channel state information(CSI). To achieve a good performance in adaptive OFDM systems, CSI should be transmitted from receiver to transmitter in real time through feedback channel. However, practically, the CSI feedback delay d which is the sum of the data processing delay and the propagation delay is not negligible and damages to the reliability of CSI such that the performance of adaptive OFDM is degraded. This paper presents an adaptive OFDM system with a multistep predictor on the frequency axis to effectively compensate the multiple feedback delays $d(\geq2)$. Via computer simulation we compare the proposed scheme and existing adaptive OFDM schemes with respect to data capacity and system performance.

• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.2
• /
• pp.63-68
• /
• 2018

Interests in face recognition area have been increasing due to diverse emerging applications. Face recognition algorithm from a two-dimensional source could be challenging in dealing with some circumstances such as face orientation, illuminance degree, face details such as with/without glasses and various expressions, like, smiling or crying. Hopfield Network capabilities have been used specially within the areas of recalling patterns, generalizations, familiarity recognitions and error corrections. Based on those abilities, a specific experimentation is conducted in this paper to apply the Redundant Parallel Hopfield Network on a face recognition problem. This new design has been experimentally confirmed and tested to be robust in any kind of practical situations.

• 한국IT서비스학회:학술대회논문집
• /
• 2003.05a
• /
• pp.436-446
• /
• 2003

기업의 업무을 지원하고, 조직의 방향을 유도하는 정보 관리시스템에 있어, 전통적인 정보기술 관리에 대한 방법이 개발자 중심의 산출물과 고립된(Standalone) 데이터 처리의 문제점을 나타내고 있다. 이로 말미암아, 고비용, 저통합성으로 인하여 변화하는 기업 내외적인 경영환경에 대처하는 기반구조 변경이 어려운 실적이고, 임기 응변적이고 벤더 중심적인 시스템 개발에 통합성과 업무와 정보시스템간의 얼라이먼트에 문제가 발생하고 있다. 이러한 문제를 해결하기 위하여 ITA를 기업에 적용하여 전사적인 경영과 정보기술간의 통합성과 변화하는 환경에 유연한 적응의 필요성이 대두되기에 이르렀다. 이에 대한 문제점을 해결할 수 있는 경영과 조직에 아키텍처 프레임워크를 구축하여 경영환경의 변화에 따른 능동적인 시스템의 변화관리를 해야 한다. 이에 따라 본 논문에서는 ZachmanFramework,IndexFramework을 소개하고 이들 Framework과 TOGAF를 비교 분석한 다음 향후 기업에 맞는 ITA를 구축하는데 있어 TOGAF의 아키텍처 프레임워크의 활용방안을 고찰하려 한다.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.249-263
• /
• 2023

Collaborative Filtering, a representative recommendation system methodology, consists of two approaches: neighbor methods and latent factor models. Among these, the latent factor model using matrix factorization decomposes the user-item interaction matrix into two lower-dimensional rectangular matrices, predicting the item's rating through the product of these matrices. Due to the factor vectors inferred from rating patterns capturing user and item characteristics, this method is superior in scalability, accuracy, and flexibility compared to neighbor-based methods. However, it has a fundamental drawback: the need to reflect the diversity of preferences of different individuals for items with no ratings. This limitation leads to repetitive and inaccurate recommendations. The Adaptive Deep Latent Factor Model (ADLFM) was developed to address this issue. This model adaptively learns the preferences for each item by using the item description, which provides a detailed summary and explanation of the item. ADLFM takes in item description as input, calculates latent vectors of the user and item, and presents a method that can reflect personal diversity using an attention score. However, due to the requirement of a dataset that includes item descriptions, the domain that can apply ADLFM is limited, resulting in generalization limitations. This study proposes a Generalized Adaptive Deep Latent Factor Recommendation Model, G-ADLFRM, to improve the limitations of ADLFM. Firstly, we use item ID, commonly used in recommendation systems, as input instead of the item description. Additionally, we apply improved deep learning model structures such as Self-Attention, Multi-head Attention, and Multi-Conv1D. We conducted experiments on various datasets with input and model structure changes. The results showed that when only the input was changed, MAE increased slightly compared to ADLFM due to accompanying information loss, resulting in decreased recommendation performance. However, the average learning speed per epoch significantly improved as the amount of information to be processed decreased. When both the input and the model structure were changed, the best-performing Multi-Conv1d structure showed similar performance to ADLFM, sufficiently counteracting the information loss caused by the input change. We conclude that G-ADLFRM is a new, lightweight, and generalizable model that maintains the performance of the existing ADLFM while enabling fast learning and inference.

• Proceedings of the KOR-KST Conference
• /
• 1998.10b
• /
• pp.191-191
• /
• 1998

ITS를 구성하는 서비스나 기능이 대부분 동적인 특성을 지니고 있어, 앞으로는 이를 효율적으로 뒷받침할 수 있는 공간데이터(Spatial Data)가 필요하다. 특히, 대부분의 ITS서비스와 기능이 정보의 신속한 전달을 위해서 유무선통신을 사용할 것이다. 또한, 최정 사용자서비스와 응용분야가 공간데이터라는 기본적인 정보를 공유하므로써 위치에 대한 정적·동적교통정보를 제공받게 된다. 정보사용자가 사용하는 공간데이터가 상이하다면, 정보의 공유가 이루어지지 않거나 정확하게 제공되지 않는 것은 자명한 사실이다. 이러한 이유는 정보사용자가 자신들의 정보수집, 정보전달, 정보분석 등의 목적에 적합한 공간데이터를 제작하여 유지하기 때문이다. 결과적으로 정보의 공유를 위해서는 상이한 공간데이터들 간에 동일한 교통정보를 공유하도록 하는 조작이나 방법이 필요하다. 서로 다른 원본으로 구성된 데이터를 통합하고 이를 ITS서비스와 기능을 위한 각 시스템에 적용하기 위해서는 서로 다른 수준을 가지고 있는 공간데이터(수치지도 데이터)의 해상도, 위치정확도, 속성정확도, 정밀도, 범위 등과 같은 문제들이 최종 응용시스템에 적용되어져야 하고, 이를 통해 공간적인 위치와 수치지도를 구성하는 각종 엔터티가 참조되어야 할 것이다. 이뿐 아니라, 향후 데이터 공유의 방법에 있어서도, 각종 무선통신의 발달과 인터넷과 같은 정보전달매체의 대중화가 이루어짐에 따라, 정보의 공유가 동시적으로 이루어질 것이다. 본 연구에서는 공공기관주도로 제작된 전국범위의 수치지도를 하여, ITS용 네트워크데이터구성을 위한 기능분석과 사양을 제시를 함으로써, 이에 대한 프로파일 개발한다. 정보공유를 위한 위치참조모델(LRM)과 프로파일을 ITS데이터에 적용함으로써, 위치참조모델의 기능과 적용성을 평가한다.키기 위한 향후의 연구과제를 제시한다.Si결정의 크기를 비교하였을 때 45$\mu\textrm{m}$ 이하의 분말을 섞어 압출하였을 때 가장 작은 초정 Si입자 크기를 얻음 을 볼 수 있었다. 주의 Fairfax County에 소재한 주간 고속도로 66번(I-66)과 인접 교통망의 교통자료를 사용하여 각종 돌발교통 혼잡 상황을 전제로 한 Traffic Simulation과 정보제공시나\리오를 INTEGRATION Model을 이용해 실행하였다. 그 결과 적응형 알고리즘이 개개인의 최단시간 경로를 제공하는 사용자 평형 경로안내전략에 비해 교통혼잡도와 정체시간의 체류정도에 따라 3%에서 10%까지 전체통행시간을 절약할 수 있다는 결론을 얻었다.출발참, 구성대외개방선면축심, 실현국제항선적함접화국내항반적전항, 형성다축심복사식항선망; 가강기장건설, 개피포동제이국제기장건설, 괄응포동개발경제발전적수요. 부화개시일은 각 5월 26일과 5월 22일이었다. 11. 6월 중순에 애벌레를 대상으로 처리한 Phenthoate EC가 96.38%의 방제가로 약효가 가장 우수하였고 3월중순 및 4월중순 월동후 암컷을 대상으로 처리한 Machine oil, Phenthoate EC 및 Trichlorfon WP는 비교적 약효가 낮았다.>$^{\circ}$E/$\leq$30$^{\circ}$NW 단열군이 연구지역 내에서 지하수 유동성이 가장 높은 단열군으로 추정된다. 이러한 사실은 3개 시추공을 대상으로 실시한 시추공 내 물리검층과 정압주입시험에서도 확인된다.. It was resulted from increase of weight of single cocoon. "Manta"2.5ppm produced 22.2kg of cocoon. It