• Journal of Information Processing Systems
• /
• 제11권1호
• /
• pp.39-56
• /
• 2015

Continuous multi-interval prediction (CMIP) is used to continuously predict the trend of a data stream based on various intervals simultaneously. The continuous integrated hierarchical temporal memory (CIHTM) network performs well in CMIP. However, it is not suitable for CMIP in real-time mode, especially when the number of prediction intervals is increased. In this paper, we propose a real-time integrated hierarchical temporal memory (RIHTM) network by introducing a new type of node, which is called a Zeta1FirstSpecializedQueueNode (ZFSQNode), for the real-time continuous multi-interval prediction (RCMIP) of data streams. The ZFSQNode is constructed by using a specialized circular queue (sQUEUE) together with the modules of original hierarchical temporal memory (HTM) nodes. By using a simple structure and the easy operation characteristics of the sQUEUE, entire prediction operations are integrated in the ZFSQNode. In particular, we employed only one ZFSQNode in each level of the RIHTM network during the prediction stage to generate different intervals of prediction results. The RIHTM network efficiently reduces the response time. Our performance evaluation showed that the RIHTM was satisfied to continuously predict the trend of data streams with multi-intervals in the real-time mode.

• 한국멀티미디어학회논문지
• /
• 제8권6호
• /
• pp.844-855
• /
• 2005

This paper proposes a human action recognition scheme for Intelligent Web Browser. Based on the principle that a human action can be defined as a combination of multiple articulation movements, the inference of stochastic grammars is applied to recognize each action. Human actions in 3 dimensional (3D) world coordinate are measured, quantized and made into two sets of 4-chain-code for xy and zy projection planes, consequently they are appropriate for applying the stochastic grammar inference method. We confirm this method by experiments, that various physical actions can be classified correctly against a set of real world 3D temporal data. The result revealed a comparatively successful achievement of $93.8\%$ recognition rate through the experiments of 8 movements of human head and $84.9\%$ recognition rate of 60 movements of human upper body. We expect that this scheme can be used for human-machine interaction commands in a web browser.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2009년도 추계학술발표대회
• /
• pp.287-288
• /
• 2009

Multi-query optimization (MQO) is a critical research issue in the real-time data stream management system (DSMS). We propose to address this problem in the ubiquitous GIS (u-GIS) environment, focusing on grouping 'similar' spatio-temporal queries incrementally into N clusters so that they can be processed virtually as N queries. By minimizing N, the overlaps in the data requirements of the raw queries can be avoided, which implies the reducing of the total disk I/O cost. In this paper, we define the spatio-temporal query clustering problem and give a data cubing approach (Q-cube), which is expected to be implemented in the cloud computing paradigm.

• 한국지구과학회지
• /
• 제27권1호
• /
• pp.83-94
• /
• 2006

이 논문에서는 2002년 9월 태풍 루사로 인해 많은 재해 피해를 입은 강원도 강릉시 사천천 유역을 대상으로 다중 시기 원격탐사 자료를 이용하여 퇴적 지질환경 변화 정보를 추출하고 분석을 수행하였다. 다중 시기 자료에 대해 자동 임계치 설정 기반 무감독 변화 탐지 기법을 적용하여 여러 시기 및 센서별 변화 정보를 추출하였다. 변화탐지 결과, 제외지에서는 태풍 루사 직후 하천 탁도 변화, 습지의 수계 혹은 퇴적물로의 변화 및 계절적인 유량 차이에 의한 하도 노출 여부 등으로 변화지역이 나타났다. 주변 농경지에서는 홍수 및 산사태 등으로 인한 토사의 퇴적, 농지 개간 등으로 인한 변화가, 기타 지역에서는 제방 공사 등으로 인한 변화가 두드러지게 나타났다. 노한 야외 조사와 원격탐사 자료를 이용하여 미지형 분류도, 범람원 지역 지표 퇴적량 분포도 및 수해 지형 분류도를 작성하였다. 결론적으로 다중시기 고해상도 원격탐사 자료가 재해로 인한 변화 정보 추출에 유용하게 활용될 수 있을 것으로 기대되며, 이를 위해 고해상도 자료에 적합한 자료처리 기법 개발이 병행되어야 할 것으로 판단된다.