• 한국컴퓨터그래픽스학회논문지
• /
• 제29권5호
• /
• pp.11-20
• /
• 2023

본 논문에서는 손 동작을 포함한 전신 동작 생성이 가능하고 동작 생성 딜레이를 조절할 수 있는 새로운 딥러닝 기반 동작 복원 기술을 제안한다. 제안된 방법은 범용적으로 사용되는 센서인 바이브 트래커와 딥러닝 기술의 융합을 통해 더욱 정교한 동작 복원을 가능하게함과 동시에 IK 솔버(Inverse Kinematics solver)를 활용하여 발 미끄러짐 현상을 효과적으로 완화한다. 본 논문은 학습된 오토인코더(AutoEncoder)를 사용하여 트래커 데이터에 적절한 캐릭터 동작의 실시간 복원이 가능하고, 동작 복원 딜레이를 조절할 수 있는 방법을 제안한다. 복원된 전신 동작에 적합한 손 동작을 생성하기 위해 FCN(Fully Connected Network)을 사용하여 손 동작을 생성하고, 오토인코더에서 복원된 전신 동작과 FCN 에서 생성된 손 동작을 합쳐 손 동작이 포함된 캐릭터의 전신 동작을 생성할 수 있다. 앞서 딥러닝 기반의 방법으로 생성된 동작에서 발 미끄러짐 현상을 완화시키기 위해 본 논문에서는 IK 솔버 를 활용한다. 캐릭터의 발에 위치한 트래커를 IK 솔버의 엔드이펙터(end-effector)로 설정하여 캐릭터의 발 움직임을 정확하게 제어하고 보정하는 기술을 제안함으로써, 생성된 동작의 전반적인 정확성을 향상시켜 고품질의 동작을 생성한다. 실험을 통해, 본 논문에서 제안한 딥러닝 기반 동작 복원에서 정확한 동작 생성과 사용자 입력에 따라 프레임 딜레이 조정이 가능함을 검증하였고, 생성된 전신 동작의 발미끄러짐 현상에 대해 IK 솔버가 적용되기 이전 전신 동작과 비교하여 보정에 대한 성능을 확인하였다.

• 한국운동역학회지
• /
• 제25권4호
• /
• pp.431-440
• /
• 2015

Objective : The purpose of this study was to investigate the relations between accuracy score and the motions which are performed in PyeongWon Poomsae, thereby developing objectivity in accuracy evaluations for Poomsae competitions. Method : The subjects were six male Poomsae players (age: $21.2{\pm}1.17yrs$, height: $173.4{\pm}3.95cm$, weight: $67.7{\pm}7.65kg$). A three-dimensional motion analysis was performed on the subjects using six high-speed cameras (60 frames/sec) and subjects' performed motions were evaluated by 5 evaluators. The entire Pyeong Poomsae was divided into 13 events and 9 phases; there were three pairs of symmetric phases among them: front kick & turning side kick phases (3PH, 3-1PH), arm motion & stance phases (4PH, 4-1PH), side kick with Hakdari-stance phases (5PH, 5-1PH). Performance time, change and range of COM, height of vertex, and foot of side kicks were analysed. The Data was analysed utilizing correlation analysis. Results : There was a positive correlation between accuracy score and the difference between right and left range of COM (X direction) at 4PH (r=0.921, p=0.009). Conclusion : The results of our study indicate that it is necessary to consider some of objective criterion such as performance time, COM range, and symmetrical movements in accuracy evaluations of Poomsae competitions.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2003년도 춘계학술대회 논문집
• /
• pp.343-346
• /
• 2003

In order to discuss the availability of the hydrostatic guideway driven by the coreless linear motor to ultra precision machine tools, a prototype of guideway is designed and tested in this research. A coreless linear DC motor with the continuous force of 156 N and a laser scale with the resulution of 10 nm are used as the feeding system. The experiments are performed on the static stiffness. motion accuracy, positioning accuracy. microstep response and variation of velocity. The guideway has the infinite axial stiffness within 50 N of applied load, and has 0.08 ${\mu}{\textrm}{m}$ of linear motion error and 0.1 arcsec of angular motion error. It also has 0.21 ${\mu}{\textrm}{m}$ of positioning error and 0.09 ${\mu}{\textrm}{m}$ of repeatability, and it shows the stable response against the 10 nm resolution step command. The velocity variation of feeding system is less than 5%. From these results, it is confirmed that the hydrostatic guideway driven by the coreless linear motor is very useful for the ultra precision machine tools.

• Geomechanics and Engineering
• /
• 제30권1호
• /
• pp.1-10
• /
• 2022

Seismic ground response evaluation is one of the main issues in geotechnical earthquake engineering. These analyses are subsequently divided into one-, two- and three-dimensional methods, and each of which can perform in time or frequency domain. In this study, a novel approach is proposed to assess the seismic site response using two-dimensional transfer functions in frequency domain analysis. Using the proposed formulation, a program is written in MATLAB environment and then promoted utilizing the equivalent linear approach. The accuracy of the written program is evaluated by comparing the obtained results with those of actual recorded data in the Gilroy region during Loma Prieta (1989) and Coyote Lake (1979) earthquakes. In order to precise comparison, acceleration time histories, Fourier amplitude spectra and acceleration response spectra diagrams of calculated and recorded data are presented. The proposed 2D transfer function diagrams are also obtained using mentioned earthquakes which show the amount of amplification or attenuation of the input motion at different frequencies while passing through the soil layer. The results of the proposed method confirm its accuracy and efficiency to evaluate ground motion during earthquakes using two-dimensional model. Then, studies on irregular topographies are carried out, and diagrams of amplification factors are shown.

• 한국정보기술학회 영문논문지
• /
• 제10권1호
• /
• pp.15-23
• /
• 2020

In this paper, we propose a new method to detect traffic accidents in video from vehicle-mounted cameras (vehicle black box). We use the distance between vehicles to determine whether an accident has occurred. To calculate the position of each vehicle, we use object detection and tracking method. By the way, in a crowded road environment, it is so difficult to decide an accident has occurred because of parked vehicles at the edge of the road. It is not easy to discriminate against accidents from non-accidents because a moving vehicle and a stopped vehicle are mixed on a regular downtown road. In this paper, we try to increase the accuracy of the vehicle accident detection by using not only the motion of the surrounding vehicle but also ego-motion as the input of the Recurrent Neural Network (RNN). We improved the accuracy of accident detection compared to the previous method.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제13권1호
• /
• pp.260-277
• /
• 2021

The appropriate radiation conditions of ship motion problem with advancing speed in frequency domain are investigated from a theoretical and practical point of view. From extensive numerical experiments that have been conducted for evaluation of the relevant radiation conditions, a hybrid radiation technique is proposed in which the Sommerfeld radiation condition and the free surface damping are mixed. Based on the comparison with the results of the translating and pulsating Green function method, the optimal damping factor of the hybrid radiation technique is selected, and the observed limitations of the proposed hybrid radiation technique are discussed, along with its accuracy obtained from the numerical solutions. Comparative studies of the forward-speed seakeeping prediction methods available confirm that the results of applying the hybrid radiation technique are relatively similar to those obtained from the translating and pulsating Green function method. This confirmation is made in comparisons with the results of solely applying either the free surface damping, or the Sommerfeld radiation condition. By applying the proposed hybrid radiation technique, the wave patterns, hydrodynamic coefficients, and motion responses of the Wigley III hull are finally calculated, and compared with those of model tests. It is found that, in comparison with the model test results, the three-dimensional Rankine source method adopting the proposed hybrid radiation technique is more robust in terms of accuracy and numerical stability, as well as in obtaining the forward speed seakeeping solution.

• 한국멀티미디어학회논문지
• /
• 제25권8호
• /
• pp.1109-1121
• /
• 2022

As the demand for quality sleep increases in modern society, the importance of sleep technology has increased. Recently, development of sleep environment improvement products and research on the user's sleep improvement have been activated. Representatively, user sleep pattern analysis research is being conducted through the existing polysomnography, but it is difficult to use it in the sleep environment of daily life. Therefore, in this paper, we propose a system design that can provide a customized deep sleep service to users by detecting sleep disturbance factors in a motion bed environment. In order to improve the user's sleep satisfaction, a logistic regression-based sleep pattern analysis model is proposed and accuracy and significance are verified through experiments. And to improve user's sleep satisfaction, we propose a logistic regression-based sleep pattern analysis model and verify accuracy and significance through experiments. The proposed system is expected to improve the user's sleep quality and effectively prevent and manage sleep disorders.

• 한국해양공학회지
• /
• 제36권5호
• /
• pp.295-302
• /
• 2022

The initial response to a marine accident can play a key role to minimize the accident. Therefore, various decision support systems have been developed using sensors, simulations, and active response equipment. In this study, we developed an algorithm to predict damage locations using ship motion data with bidirectional long short-term memory (BiLSTM), a type of recurrent neural network. To reflect the low frequency ship motion characteristics, 200 time-series data collected for 100 s were considered as input values. Heave, roll, and pitch were used as features for the prediction model. The F1-score of the BiLSTM model was 0.92; this was an improvement over the F1-score of 0.90 of a prior model. Furthermore, 53 of 75 locations of damage had an F1-score above 0.90. The model predicted the damage location with high accuracy, allowing for a quick initial response even if the ship did not have flood sensors. The model can be used as input data with high accuracy for a real-time progressive flooding simulator on board.

• 국제학술발표논문집
• /
• The 8th International Conference on Construction Engineering and Project Management
• /
• pp.443-452
• /
• 2020

Worker's awkward postures and unreasonable physical load can be corrected by monitoring construction activities, thereby increasing the safety and productivity of construction workers and projects. However, manual identification is time-consuming and contains high human variance. In this regard, an automated activity recognition system based on inertial measurement unit can help in rapidly and precisely collecting motion data. With the acceleration data, the machine learning algorithm will be used to train classifiers for automatically categorizing activities. However, input acceleration data are extracted either from designed experiments or simple construction work in previous studies. Thus, collected data series are discontinuous and activity categories are insufficient for real construction circumstances. This study aims to collect acceleration data during long-term continuous work in a construction project and validate the feasibility of activity recognition algorithm with the continuous motion data. The data collection covers two different workers performing formwork at the same site. An accelerator, as well as portable camera, is attached to the worker during the entire working session for simultaneously recording motion data and working activity. The supervised machine learning-based models are trained to classify activity in hierarchical levels, which reaches a 96.9% testing accuracy of recognizing rest and work and 85.6% testing accuracy of identifying stationary, traveling, and rebar installation actions.

• 한국도로학회논문집
• /
• 제12권4호
• /
• pp.187-196
• /
• 2010

본 논문의 목적은 장기적인 차량중량자료의 획득을 위하여 설치된 Weigh-In-Motion(WIM)시스템의 정확도를 개선하기 위한 온도영향 보정기법을 개발하는 것이다. 이를 위해 국내 교통량통계연보에 근거하여 중차량교통량이 많은 지역 중 한 곳을 선정하고 WIM 시스템을 설치하여 2010년 1월부터 차량의 중량자료를 획득하였다. 본 연구에서는 획득한 자료 가운데 5월부터 8월까지의 WIM 자료에 대하여 온도에 따른 영향을 보정하기 위하여 화물적재량에 따른 영향을 가장 덜 받아 상대적으로 일정한 중량값을 나타내는 10종 차량(5축 세미트랙터트레일러)의 첫 번째 축의 중량자료를 이용하였다. 이로부터 일평균, 최대 및 최저기온에 따른 보정식을 개발하였다. 마지막으로 새로 얻어진 차량중량자료를 분석하여 보정방법의 적정함을 나타내었고 재현주기별 극한하중 예측을 통해 보정 전,후의 자료값의 차이를 비교하였다. 이 연구의 결과는 WIM system으로 얻게 되는 장기적인 차량중량자료의 정확도를 향상시킴으로써 도로포장 설계 및 교량설계를 위한 설계차량활하중모형의 합리적인 선택에 기여할 것으로 판단된다.