• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권10호
• /
• pp.4117-4135
• /
• 2020

Due to the angle of view, illumination and scene diversity, real-time dynamic face detection and recognition is no small difficulty in those unrestricted environments. In this study, we used the intrinsic correlation between detection and calibration, using a multi-task cascaded convolutional neural network(MTCNN) to improve the efficiency of face recognition, and the output of each core network is mapped in parallel to a compact Euclidean space, where distance represents the similarity of facial features, so that the target face can be identified as quickly as possible, without waiting for all network iteration calculations to complete the recognition results. And after the angle of the target face and the illumination change, the correlation between the recognition results can be well obtained. In the actual application scenario, we use a multi-camera real-time monitoring system to perform face matching and recognition using successive frames acquired from different angles. The effectiveness of the method was verified by several real-time monitoring experiments, and good results were obtained.

• The Journal of Korean Physical Therapy
• /
• 제33권3호
• /
• pp.136-141
• /
• 2021

Purpose: Dementia is a disease in which cognitive function declines, leading to deterioration of body functions and activities of daily living. The purpose of this study is to explore the effects of dual-task training, including cognitive tasks, on cognitive and body function and β-amyloid levels in Alzheimer's dementia patients. Methods: 34 inpatients diagnosed with Alzheimer's dementia at a nursing hospital located in South Korea. The patients were randomly divided into a dual-task group (n=16) and a single-task group (n=18). Each group was trained for 30 minutes three times a week for eight weeks. The MMSE-K was used to measure the patients' cognitive function. To assess the patients' static balance ability, their LOS was measured using BioRescue. while dynamic balance was measured using the BBS. The 10MWT were conducted to evaluate the patients' walking ability. Blood analysis was performed to measure levels of β-amyloid. Results: Both groups exhibited statistically significant improvements in gait function after the training (p<0.05). The dual-task group exhibited statistically significant differences in cognitive function, static and dynamic balance function, and β-amyloid levels after training (p<0.05). A significant difference was observed between the two groups (p<0.05). Conclusion: Dual-task training were found to be effective in improving cognitive and bodily functioning and reducing β-amyloid levels in Alzheimer's dementia patients. Thus, this may be suggested as an effective exercise method for the treatment and early prevention of Alzheimer's dementia.

• 제어로봇시스템학회논문지
• /
• 제6권1호
• /
• pp.66-72
• /
• 2000

As the demand for the design of modular manipulators or special purpose manipulators has increased, task based design to design an optimal manipulator for a given task become more and more important. However, the complexity with a large number of design parameters, and highly nonlinear and implicit functions are characteristics of a general manipulator design. To achieve the goal of task based design, it is necessary to develop a methodology to solve the complexity. This paper addresses how to determine the kinematic parameters of a two-degrees of freedom manipulator with parallelogram five-bar link mechanism from a given task, namely, how to map a given task into the kinematic parameters. With simplified example of designing a manipulator with five-bar link mechanism, the methodology for task based design is presented. And it introduces formulations of a given task and manipulator specifications, and presents a new dexterity measure for manipulator design. Also the optimization problem with constraints is solved by using a genetic algorithm that provides robust search in complex spaces.

• Journal of Information Processing Systems
• /
• 제17권5호
• /
• pp.905-917
• /
• 2021

In recent years, edge computing technology consists of several Internet of Things (IoT) devices with embedded sensors that have improved significantly for monitoring, detection, and management in an environment where big data is commercialized. The main focus of edge computing is data optimization or task offloading due to data and task-intensive application development. However, existing offloading approaches do not consider correlations and associations between data and tasks involving edge computing. The extent of collaborative offloading segmented without considering the interaction between data and task can lead to data loss and delays when moving from edge to edge. This article proposes a range segmentation of dynamic offloading (RSDO) algorithm that isolates the offload range and collaborative edge node around the edge node function to address the offloading issue.The RSDO algorithm groups highly correlated data and tasks according to the cause of the overload and dynamically distributes offloading ranges according to the state of cooperating nodes. The segmentation improves the overall performance of edge nodes, balances edge computing, and solves data loss and average latency.

• 대한인간공학회:학술대회논문집
• /
• 대한인간공학회 1998년도 춘계학술대회논문집
• /
• pp.194-199
• /
• 1998

The purpose of this study is conducted to evaluate the psycho-physiological effects of the dynamic lighting on VDT workers and to establish the dynamic lighting design guides. The main experimental factors are changes of illuminance. There were five conditions including 1) ripid change, 2) slow change at 300lx .approx. 500lx, 3) a static illuminance level at 300lx, and 4) rapid chante, 5) slow change at 200lx .approx. 300lx. The ranges of change were set typical illuminance level for the simple visual task requirement and reasonable to a VDT task ; 200-300-500lx. The psycho-physiological measurements used were EEG, R-R interval, CFF, near point accommodation, feeling perceived fatigue, and work performance. Analysis of psycho-physiological factors shows that higher illumination level conditions are more affirmative effects than lower on VDT workers under 500lx except for the results of feeling perceived fatigue. And about the speed of changes of illuminance level. The conditions of slow change were better than the others in terms of the work performance as well as psycho-physiological results.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2006년도 하계종합학술대회
• /
• pp.925-926
• /
• 2006

In the multi-agent system for a single task, the action selection can be made for the real-time environment by using the global coordination space, global coordination graph and MAX-PLUS algorithm. However, there are some difficulties in multi-agent system for multi-tasking. In this paper, a real-time decision making method is suggested by using coordination space, coordination graph and dynamic coordinated state of multi-agent system including many agents and multiple tasks. Specifically, we propose locally dynamic coordinated state to effectively use MAX-PLUS algorithm for multiple tasks completion. Our technique is shown to be valid in the box pushing simulation of a multi-agent system.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1997년도 한국자동제어학술회의논문집; 한국전력공사 서울연수원; 17-18 Oct. 1997
• /
• pp.1293-1296
• /
• 1997

We investigate a design methodology for master arm which is employed in remote-site control. Though counter-balancing has been employed for the purpose of lessening the actuator loads, it results in the increase of dynamic load at the task position. Therefore, there exist a trade-off between counter-balacing and dynamic performance. The concept of a composite index is introduced to simulataneously consider those two effects in the design of master arms. Several alternative designs of master arms are suggested.

• International Journal of Control, Automation, and Systems
• /
• 제4권2호
• /
• pp.197-203
• /
• 2006

This paper addresses dynamic modeling and task-space trajectory following issues for nonholonomic mobile manipulators moving on a slope. An integrated dynamic modeling method is proposed considering nonholonomic constraints and interactive motions. An adaptive neural-fuzzy controller is presented for end-effector trajectory following, which does not rely on precise apriori knowledge of dynamic parameters and can suppress bounded external disturbances. Effectiveness of the proposed algorithm is verified through simulations.

• 대한물리의학회지
• /
• 제11권2호
• /
• pp.83-91
• /
• 2016

PURPOSE: The purpose of this study was to determine the efficacy of dual-task action observation training (AOT) and single-task AOT related with daily living task on walking ability and ADL performance in chronic stroke patients. METHODS: Twenty-seven chronic stroke subjects were included in the study. They were randomly assigned to three task categorieds as follows: whole dual-task AOT or partial dual-task AOT or single-task AOT rehabilitation. Whole dual-task AOT observed the movement at once and partial dual-task AOT observed the movement divided into 4 parts related functional gait and activities of daily living task for 2 minutes 30 seconds. Single-task AOT observed the movement related functional gait for 2 minutes 30 seconds. Both groups had physical training session for 12 minutes 30 seconds. The study was conducted for four weeks, with three training sessions a week, for twelve weeks. All subjects were evaluated for their walking ability and activities of daily living through devices, 10m walking test (10MWT), dynamic gait index (DGI), and Korea-Modified Barthel Index (K-MBI). RESULTS: A significant improvement of walking ability and ADL performance happened among dual-task AOT subjects, compared with a single-task AOT subjects, during the 4-weeks course treatment. The results of the study showed statistically significant differences in 10MWT (p<0.05) and DGI (p<0.05), and K-MBI (p<0.05). CONCLUSION: Our results indicated that dual-task AOT has a positive additional impact on recovery of walking ability and ADL performance in chronic stroke patients.

• 한국콘텐츠학회논문지
• /
• 제7권1호
• /
• pp.83-93
• /
• 2007

본 논문에서는 주기적으로 생성(release)되는 주기성 태스크(Periodic Task)들과 산발적으로 생성되는 산발성 태스크(Sporadic Task)들이 혼합된 실시간 시스템을 위한 전력관리 스케줄링 기법을 제안한다. 각각의 태스크는 최소주기, 최악수행요구시간과 마감시간 등으로 정의된다. 본 논문에서 제안한 동적 전압조정(Dynamic Voltage Scaling : DVS) 알고리즘인 DVSMT(DVS for mixed tasks)는 태스크의 실시간 마감시간을 보장하면서 작업이 종료됐을 때, 수행하는 동안 사용한 사이클 중 다른 태스크들이 할당한 수행 사이클을 자신의 마감시간까지 온라인 상태에서 균등 분배함으로써 공급전압(또한 동작 주파수)을 동적으로 조정한다. 이러한 기법으로 더 많은 에너지를 절감할 수 있다. 제안한 알고리즘은 실시간 운영체제에 쉽게 통합될 수 있기 때문에 제한된 배터리 전력을 이용하는 휴대용 기기 및 센서망 노드 등에 적용할 수 있다. 시뮬레이션 결과들은 DVSMT가 주기성 태스크들로만 구성된 시스템과 주기성 태스크들 및 산발성 태스크들이 혼합된 시스템에서 기존의 알고리즘보다 대략 60% 까지 에너지가 절감됨을 보였다.