• Korean Journal of Radiology
• /
• 제25권3호
• /
• pp.224-242
• /
• 2024

The emergence of Chat Generative Pre-trained Transformer (ChatGPT), a chatbot developed by OpenAI, has garnered interest in the application of generative artificial intelligence (AI) models in the medical field. This review summarizes different generative AI models and their potential applications in the field of medicine and explores the evolving landscape of Generative Adversarial Networks and diffusion models since the introduction of generative AI models. These models have made valuable contributions to the field of radiology. Furthermore, this review also explores the significance of synthetic data in addressing privacy concerns and augmenting data diversity and quality within the medical domain, in addition to emphasizing the role of inversion in the investigation of generative models and outlining an approach to replicate this process. We provide an overview of Large Language Models, such as GPTs and bidirectional encoder representations (BERTs), that focus on prominent representatives and discuss recent initiatives involving language-vision models in radiology, including innovative large language and vision assistant for biomedicine (LLaVa-Med), to illustrate their practical application. This comprehensive review offers insights into the wide-ranging applications of generative AI models in clinical research and emphasizes their transformative potential.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2006년도 영호남 합동 학술대회 및 춘계학술대회 논문집 센서 박막 기술교육
• /
• pp.15-19
• /
• 2006

A pulsed laser ablation deposition (PLAD) technique is an excellent method for the fabrication of amorphous carbon (a-C) films, because it can generate highly energetic carbon clusters on a substrate. This paper was focused on the understanding and analysis of the motion of C particles in laser ablation assisted by Ar plasmas. The simulation has carried out under the pressure P=50 mTorr of Ar plasmas. Two-dimensional hybrid model consisting of fluid and Monte-Carlo models was developed and three kinds of the ablated particles which are carbon atom (C), ion ($C^+$) and electron were considered in the calculation of particle method. The motions of energetic $C^+$ and C deposited upon the substrate were investigated and compared. The interactions between the ablated particles and Ar gas plasmas were discussed.

• 한국음향학회지
• /
• 제13권1호
• /
• pp.24-31
• /
• 1994

본 논문에서는 연속분포 hidden Markov모델을 이용한 화자독립 연속 음성 인식 시스템에 관해 기술한다. 연속분포 모델은 평균과 분산 벡터로 구성되며 음성신호를 직접 모델링하여 양자화 왜곡이 없어진다. 특징벡터는 filter bank 계수 및 그 1, 2차 미분계수를 사용하여 음성신호의 동적 특성을 반영하였다. Segmental K-means 알고리즘을 이용하여 학습하였으며, 연속어 인식에서 가장 문제가 되는 조음화 현상으로 인한 인식률 저하를 막기 위해 앞뒤의 음소를 고려해주는 triphone을 인식단위로 사용하였다. Search 알고리즘으로는 시간 면에서 효율이 좋은 one-pass search 알고리즘을 사용하였다 성능 평가를 위한 회자 독립인식 실험에서 문법이 없을 경우 $83\%$, finite state network을 적용한 경우에는 $94\%$의 인식률을 나타내었다.

• 대한기계학회논문집A
• /
• 제20권1호
• /
• pp.146-153
• /
• 1996

A SCARA type direct drive robot which can be used in the assembly operation was designed and manufactured. Graphite fiber epoxy composite material was used in the fabrication of the robot arm structure in order to improve the speed of the robot arm with a high damping effect. For model-based control and sensitivity analysis of system parameters, the dynamic model of robot arm and drive servo amplifier parameters such as equivalent gains of PWM driver and velocity gains of servo system were estimated from frequency response tests. The complete dynamic model for overall robot system was used in the simulation of the open-loop control. The simulation results agreed reasonably well to the experimental results. The feedforward control using the dynamic models improved the trajectory tracking performance, decreasing the tracking error by factor of three compared with PID control. This study found that the inverse dynamic model of the robot arm including the drive servo system showed better performances than the case of arm dynamic model only.

• 한국방사성폐기물학회:학술대회논문집
• /
• 한국방사성폐기물학회 2003년도 가을 학술논문집
• /
• pp.534-538
• /
• 2003

Final disposal of radioactive waste generated from Nuclear Power Plant (NPP) requires the detailed knowledge of the natures and quantities of radionuclides in waste package. Many of these radionuclides are difficult to measure and expensive to assay. Thus it is suggested to the Indirect method by which the concentrations of DTM (Difficult-to-Measure) nuclide is decided using the relation of concentrations (Scaling Factor) between Key (Easy-to-Measure) nuclide and DTM nuclide with measured concentrations of Key nuclide. In general, scaling factor is determined by using of log mean average (LMA) and regression. These methods are adequate to apply most corrosion product nuclides. But in case of fission product nuclides and some corrosion product nuclides, the predicted values aren't well matched with the original values. In this study, the models using artificial neural network (ANN) for C-14 and Sr-90 are compared with those using LMA and regression. The assessment of models is executed in the two parts divided by a training part and a validation part. For all of two nuclides in the training part, the predicted values using ANN are well matched with the measured values compared with those using LMA and regression. In the validation part, the accuracy of the predicted values using ANN is better than that using LMA and is similar to or better than that using regression. It is concluded that the predicted values using ANN model are better than those using conventional model in some nuclides and ANN model can be used as the complement of LMA and regression model.

• Nuclear Engineering and Technology
• /
• 제49권4호
• /
• pp.721-733
• /
• 2017

When temporary spent fuel storage pools at nuclear power plants reach their capacity limit, the spent fuel must be moved to an alternative storage facility. However, radioactive materials must be handled and stored carefully to avoid severe consequences to the environment. In this study, the risks of three potential accident scenarios (i.e., maritime transportation, an aircraft crashing into an interim storage facility, and on-site transportation) associated with the spent fuel transportation process were analyzed using a probabilistic approach. For each scenario, the probabilities and the consequences were calculated separately to assess the risks: the probabilities were calculated using existing data and statistical models, and the consequences were calculated using computation models. Risk assessment software was developed to conveniently integrate the three scenarios. The risks were analyzed using the developed software according to the shipment route, building characteristics, and spent fuel handling environment. As a result of the risk analysis with varying accident conditions, transportation and storage strategies with relatively low risk were developed for regulators and licensees. The focus of this study was the risk assessment methodology; however, the applied model and input data have some uncertainties. Further research to reduce these uncertainties will improve the accuracy of this model.

• 대한조선학회논문집
• /
• 제61권3호
• /
• pp.143-151
• /
• 2024

In this study, we present a mapping framework for 3D spatial reconstruction of digital twin model using navigation and perception sensors mounted on an Autonomous Surface Vehicle (ASV). For improving the level of realism of digital twin models, 3D spatial information should be reconstructed as a digitalized spatial model and integrated with the components and system models of the ASV. In particular, for the 3D spatial reconstruction, color and 3D point cloud data which acquired from a camera and a LiDAR sensors corresponding to the navigation information at the specific time are required to map without minimizing the noise. To ensure clear and accurate reconstruction of the acquired data in the proposed mapping framework, a image preprocessing was designed to enhance the brightness of low-light images, and a preprocessing for 3D point cloud data was included to filter out unnecessary data. Subsequently, a point matching process between consecutive 3D point cloud data was conducted using the Generalized Iterative Closest Point (G-ICP) approach, and the color information was mapped with the matched 3D point cloud data. The feasibility of the proposed mapping framework was validated through a field data set acquired from field experiments in a inland water environment, and its results were described.

• Advances in aircraft and spacecraft science
• /
• 제1권3호
• /
• pp.291-310
• /
• 2014

An advanced model for the linear flutter analysis is introduced in this paper. Higher-order beam structural models are developed by using the Carrera Unified Formulation, which allows for the straightforward implementation of arbitrarily rich displacement fields without the need of a-priori kinematic assumptions. The strong form of the principle of virtual displacements is used to obtain the equations of motion and the natural boundary conditions for beams in free vibration. An exact dynamic stiffness matrix is then developed by relating the amplitudes of harmonically varying loads to those of the responses. The resulting dynamic stiffness matrix is used with particular reference to the Wittrick-Williams algorithm to carry out free vibration analyses. According to the doublet lattice method, the natural mode shapes are subsequently used as generalized motions for the generation of the unsteady aerodynamic generalized forces. Finally, the g-method is used to conduct flutter analyses of both isotropic and laminated composite lifting surfaces. The obtained results perfectly match those from 1D and 2D finite elements and those from experimental analyses. It can be stated that refined beam models are compulsory to deal with the flutter analysis of wing models whereas classical and lower-order models (up to the second-order) are not able to detect those flutter conditions that are characterized by bending-torsion couplings.

• Advanced Composite Materials
• /
• 제17권4호
• /
• pp.373-407
• /
• 2008

This paper will examine the possibilities of the virtual tests of composite structures by simulating mechanical behaviors by using supercomputing technologies, which have now become easily available and powerful but relatively inexpensive. We will describe mainly the applications of large-scale finite element analysis using the direct numerical simulation (DNS), which describes composite material properties considering individual constituent properties. DNS approach is based on the full microscopic concepts, which can provide detailed information about the local interaction between the constituents and micro-failure mechanisms by separate modeling of each constituent. Various composite materials such as metal matrix composites (MMCs), active fiber composites (AFCs), boron/epoxy cross-ply laminates and 3-D orthogonal woven composites are selected as verification examples of DNS. The effective elastic moduli and impact structural characteristics of the composites are determined using the DNS models. These DNS models can also give the global and local information about deformations and influences of high local in-plane and interlaminar stresses induced by transverse impact loading at a microscopic level inside the materials. Furthermore, the multi-scale models based on DNS concepts considering microscopic and macroscopic structures simultaneously are also developed and a numerical low-velocity impact simulation is performed using these multi-scale DNS models. Through these various applications of DNS models, it can be shown that the DNS approach can provide insights of various structural behaviors of composite structures.

• 대한금속재료학회지
• /
• 제49권6호
• /
• pp.425-439
• /
• 2011

A review of the development history of interatomic potential models for ionic materials was carried out paying attention to the way of future development of an interatomic potential model that can cover ionic, covalent and metallic bonding materials simultaneously. Earlier pair potential models based on fixed point charges with and without considering the electronic polarization effect were found to satisfactorily describe the fundamental physical properties of crystalline oxides (Ti oxides, $SiO_2$, for example) and their polymorphs, However, pair potential models are limited in dealing with pure elements such as Ti or Si. Another limitation of the fixed point charge model is that it cannot describe the charge variation on individual atoms depending on the local atomic environment. Those limitations lead to the development of many-body potential models(EAM or Tersoff), a charge equilibration (Qeq) model, and a combination of a many-body potential model and the Qeq model. EAM+Qeq can be applied to metal oxides, while Tersoff+Qeq can be applied to Si oxides. As a means to describe reactions between Si oxides and metallic elements, the combination of 2NN MEAM that can describe both covalent and metallic elements and the Qeq model is proposed.