• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2008년도 학술대회 1부
• /
• pp.487-492
• /
• 2008

Facial expression recognition is an intensive research area for designing Human Computer Interfaces. In this work, we present a new facial expression recognition system utilizing Enhanced Independent Component Analysis (EICA) for feature extraction and discrete Hidden Markov Model (HMM) for recognition. Our proposed approach for the first time deals with sequential images of emotion-specific facial data analyzed with EICA and recognized with HMM. Performance of our proposed system has been compared to the conventional approaches where Principal and Independent Component Analysis are utilized for feature extraction. Our preliminary results show that our proposed algorithm produces improved recognition rates in comparison to previous works.

• 한국컴퓨터정보학회논문지
• /
• 제23권12호
• /
• pp.195-201
• /
• 2018

This study is attempted to demonstrate the structure of the influencing factors on the performance of services of knowledge based service firms. In the model presented, the structure is that service performance in knowledge-based services are affected by the strategic utilization of knowledge resources and service orientation factors of the service organization. And the service performance are also affected in which the systemization of service processes and customer satisfaction play a mediating role. As an analysis result of examining 148 practitioners engaged in the knowledge service industry, it is necessary to increase the satisfaction of external customers in order to improve service performance in a knowledge-based service organizations. This can be achieved by increasing the satisfaction of internal customers. In addition, for this structure to be successful, the service production process to be provided to the customer must be enhanced. The service production process has been found to be an important factor influencing the internal customer satisfaction and service delivery process, especially in the use of knowledge resource, and customer oriented service among service orientation factors.

• 한국정보처리학회논문지
• /
• 제7권1호
• /
• pp.74-87
• /
• 2000

본 논문에서는 초고속 통신망인 Myrinet을 기반으로 하는 워크스테이션 연동(network of workstations) 환경하에서 병렬 프로그래밍 모델을 지원하는 PVM (parallel virtual machine)의 통신 성능 개선 방안을 제시한다. PVM을 위해 새로이 제안하는 Myrinet 기반 통신 모델은 커널을 경유하는 기존의 UDP/IP 프로토콜을 이용하는 통신 모델과 Myrinet API를 직접 이용하는 통신 모델을 혼합한 복합 통신 모델이다. 제안된 복합 통신 모델은 사용자 영역(커널)에서 커널(사용자) 버퍼로의 메시지 복사 부하를 제거하고 커널내의 프로토콜 스택 처리를 위한 통신 지연 요인을 감소시키므로, Myrinet상에서 보다 빠른 PVM 테스크간의 데이터 전송 속도를 보장한다. 또한, UDP/IP를 사용하는 기존 PVM에 본 논문에서 제안된 Myrinet 기반 복합 통신 모델을 적용시킨 EPVM (Enhanced PVM)을 구현하고, 이의 성능을 측정하였다. 실험 결과 EPVM의 통신 성능이 기존 PVM 보다 평균 1.5배 정보 우수하다는 것을 확인하였다.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2008년도 추계학술대회논문집
• /
• pp.230-236
• /
• 2008

In this paper, a model-based stick-slip compensation for the micro-positioning is proposed using an enhanced stick-slip model based on statistical rough surface contact model. The smart structure is comprised with PZT (lead (Pb) zirconia (Zr) Titanate (Ti)) based stack actuator incorporating with the PID (Proportional-Integral-Derivative) control algorithm, mechanical displacement amplifier and positioning devices. For the stick-slip compensation, the elastic-plastic static friction model is used considering the elastic-plastic asperity contact in the rough surfaces statistically. Mathematical model of system for the positioning apparatus was derived from the dynamic behaviors of structural parts. PID feedback control algorithms with the developed stick-slip model as well as feedforward friction compensator are formulated for achieving the accurate positioning performance. Experimental results are provided to show the performances of friction control using the developed positioning apparatus.

• 한국소음진동공학회논문집
• /
• 제18권11호
• /
• pp.1134-1142
• /
• 2008

In this paper, a model-based stick-slip compensation for the micro-positioning is proposed using an enhanced stick-slip model based on statistical rough surface contact model. The smart structure is comprised with PZT(lead (Pb) zirconia(Zr) Titanate(Ti)) based stack actuator incorporating with the PID(proportional-integral-derivative) control algorithm, mechanical displacement amplifier and positioning devices. For the stick-slip compensation, the elastic-plastic static friction model is used considering the elastic-plastic asperity contact in the rough surfaces statistically. Mathematical model of system for the positioning apparatus was derived from the dynamic behaviors of structural parts. PID feedback control algorithms with the developed stick-slip model as well as feedforward friction compensator are formulated for achieving the accurate positioning performance. Experimental results are provided to show the performances of friction control using the developed positioning apparatus.

• International Journal of Aeronautical and Space Sciences
• /
• 제16권4호
• /
• pp.571-580
• /
• 2015

In this paper, a computational study was conducted in order to investigate the rotor blade sweep effect on the aerodynamics of a small axial supersonic impulse turbine stage. For this purpose, three-dimensional unsteady RANS simulations have been performed with three different rotor blade sweep angles ($-15^{\circ}$, $0^{\circ}$, $+15^{\circ}$) and the results were compared with each other. Both NTG (No tip gap) and WTG (With tip gap) models were applied to examine the effect on tip leakage flow. As a result of the simulation, the positive sweep model ($+15^{\circ}$) showed better performance in relative flow angle, Mach number distribution, entropy rise, and tip leakage mass flow rate compared with no sweep model. With the blade static pressure distribution result, the positive sweep model showed that hub and tip loading was increased and midspan loading was reduced compared with no sweep model while the negative sweep model ($-15^{\circ}$) showed the opposite result. The positive sweep model also showed a good aerodynamic performance around the hub region compared with other models. Overall, the positive sweep angle enhanced the turbine efficiency.

• 한국분무공학회지
• /
• 제27권2호
• /
• pp.57-65
• /
• 2022

The purpose of this study is to use machine learning to build a model capable of predicting the flash boiling spray characteristics. In this study, the flash boiling spray was visualized using Shadowgraph visualization technology, and then the spray image was processed with MATLAB to obtain quantitative data of spray characteristics. The experimental conditions were used as input, and the spray characteristics were used as output to train the machine learning model. For the machine learning model, the XGB (extreme gradient boosting) algorithm was used. Finally, the performance of machine learning model was evaluated using R² and RMSE (root mean square error). In order to have enough data to train the machine learning model, this study used 12 injectors with different design parameters, and set various fuel temperatures and ambient pressures, resulting in about 12,000 data. By comparing the performance of the model with different amounts of training data, it was found that the number of training data must reach at least 7,000 before the model can show optimal performance. The model showed different prediction performances for different spray characteristics. Compared with the upstream spray angle and the downstream spray angle, the model had the best prediction performance for the spray tip penetration. In addition, the prediction performance of the model showed a relatively poor trend in the initial stage of injection and the final stage of injection. The model performance is expired to be further enhanced by optimizing the hyper-parameters input into the model.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제6권2호
• /
• pp.173-177
• /
• 2006

This paper investigates the asymmetric costs of false errors to enhance the detection systems performance. The proposed method utilizes the network model to consider the cost ratio of false errors. By comparing false positive errors with false negative errors this scheme achieved better performance on the view point of both security and system performance objectives. The results of our empirical experiment show that the network model provides high accuracy in detection. In addition, the simulation results show that effectiveness of probe detection is enhanced by considering the costs of false errors.

• 한국콘크리트학회:학술대회논문집
• /
• 한국콘크리트학회 2005년도 봄학술 발표회 논문집(I)
• /
• pp.207-210
• /
• 2005

The objective of this research is to evaluate of seismic performance for reinforced concrete bridge piers with lap splices of longitudinal reinforcement steels using predict of nonlinear hysteric behavior. For the purpose, analytical trilinear hysteretic model has been used to simulate the force displacement hysteretic curve of RC bridge piers under repeated reversal loads. The moment capacity and corresponding curvature in the plastic hinge have been determined, and the enhanced hysteretic behavior model by five different kinds of branches has been proposed for modeling the stiffness variation of RC section under cyclic loading. The strength and stiffness degradation index are introduced to compute the hysteretic curve vary confinement steel ratio. In addition, the modified curvature factor has been introduced to forecast of seismic performance of longitudinal steel lap spliced and retrofitted specimens.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2024년도 춘계학술발표대회
• /
• pp.692-695
• /
• 2024

With the recent advancements in artificial intelligence (AI), the performance of deep learning-based audio deepfake technology has significantly improved. This technology has been exploited for criminal activities, leading to various cases of victimization. To prevent such illicit outcomes, this paper proposes a deep learning-based audio deepfake detection model. In this study, we propose CoNSIST, an improved audio deepfake detection model, which incorporates three additional components into the graph-based end-to-end model AASIST: (i) Squeeze and Excitation, (ii) Positional Encoding, and (iii) Reformulated HS-GAL, This incorporation is expected to enable more effective feature extraction, elimination of unnecessary operations, and consideration of more diverse information, thereby improving the performance of the original AASIST. The results of multiple experiments indicate that CoNSIST has enhanced the performance of audio deepfake detection compared to existing models.