• ETRI Journal
• /
• v.44 no.2
• /
• pp.327-338
• /
• 2022

Classification models for human action recognition require robust features and large training sets for good generalization. However, data augmentation methods are employed for imbalanced training sets to achieve higher accuracy. These samples generated using data augmentation only reflect existing samples within the training set, their feature representations are less diverse and hence, contribute to less precise classification. This paper presents new data augmentation and action representation approaches to grow training sets. The proposed approach is based on two fundamental concepts: virtual video generation for augmentation and representation of the action videos through robust features. Virtual videos are generated from the motion history templates of action videos, which are convolved using a convolutional neural network, to generate deep features. Furthermore, by observing an objective function of the genetic algorithm, the spatiotemporal features of different samples are combined, to generate the representations of the virtual videos and then classified through an extreme learning machine classifier on MuHAVi-Uncut, iXMAS, and IAVID-1 datasets.

• Journal of Science of Art and Design
• /
• v.12
• /
• pp.125-143
• /
• 2007

The politics penetrating through the contemporary art since modernism to postmodernism is to accomplish the 'Non-representation' in the artworks. This study argues that postmodernism did not put an end to the formalistic feature of modernism but intended to accomplish it. Modernist art aimed at purity, i.e. self-referential and self-definition art advocated by Clement Greenberg, and it carne to the end by accomplishing flatness and materiality. It was an 'evasion to the matter' which allocated the object from visuality of outer object to the psychic image of the subject. It failed being 'non-representational' as what it really achieved was transition of object. Jean Baudrillard's theory tried to overcome the representational quality by 'being simulacre'. In the representative artworks of the past, the meaning of artworks was reverted under the outer context or object. The meaning again failed being 'Non-representational' as it was restored to the psychic image of the subject in modernist artworks where the definite illusion was demolished Meanwhile, artwork advocating simulacre acquired Non-representational quality by liberating from both models. It did not deconstruct the self-referential tendency of modernism but maximized the Non-representational modernistic principle. After creating 'Non-representation' through simulacre, the existential status and function of an artwork is the inclination and moral of contemporary art as 'Non-representation'. The image theory of Henri Bergson sets the existential status of 'image' as it does not belong to either subject nor object. It provides significant foundation for arguing the existential status of simulacre. Moreover, though an artwork as a fragment forming a movement image in the world cannot represent the object, it can however sustain certain kind of fractal resemblance with the world by letting the two parties communicate. The theory of sense by Gilles Deleuze is of profound significance as it specifically indicated way how the stage of absorption through the unity of subject and object is realized in forms of artworks, and configured the latent and invisible energy.

• Journal of the Korean School Mathematics Society
• /
• v.24 no.2
• /
• pp.217-241
• /
• 2021

The purpose of this study is to qualitatively examine the effects of graphics representation of trigonometry modelling concerning question generating and idea sharing. The experimental setting(Experiment Group) was one class (N=26) at a public high school. The modelling process was designed as a process-oriented conceptualization divided into three steps i.e., (1) game with idea sharing and question generating, (2) graphic representation, and (3) symbolization in the mathematical applied tasks related to trigonometry function. The result indicates that Graphic Representation with Game Activity increases the opportunity of question generating and idea sharing during experimental work. Also, the results show that the introduction of computer graphics enhances the teaching of mathematical quantity in highschool classrooms.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1994.04a
• /
• pp.561-568
• /
• 1994

A man model can be used as an effective tool to design ergonomically sound products and workplaces, and subsequently evaluate them properly. For a man model to be truly useful, it must be integrated with a posture prediction model which should be capable of representing the human arm reach posture in the context of equipments and workspaces. Since the human movement possesses redundant degrees of freedom, accurate representation or prediction of human movement was known to be a difficult problem. To solve this redundancy problem, a psychophysical cost function was suggested in this study which defines a cost value for each joint movement angle. The psychophysical cost function developed integrates the psychophysical discomfort of joints and the joint range availability concept which has been used for redundant arm manipulation in robotics to predict the arm reach posture. To properly predict an arm reach posture, an arm reach posture prediction model was then developed in which a posture configuration that provides the minimum total cost is chosen. The predictivity of the psychophysical cost function was compared with that of the biomechanical cost function which is based on the minimization of joint torque. Here, the human body is regarded as a two-dimensional multi-link system which consists of four links ; trunk, upper arm, lower arm and hand. Real reach postures were photographed from the subjects and were compared to the postures predicted by the model. Results showed that the postures predicted by the psychophysical cost function closely simulated human reach postures and the predictivity was more accurate than that by the biomechanical cost function.

• Korean Journal of Computational Design and Engineering
• /
• v.12 no.3
• /
• pp.220-232
• /
• 2007

Current geometric modeling and analysis are commonly based on B-Rep modeling and a finite elements method respectively. Furthermore, it is difficult to represent an object whose material property is heterogeneous using the B-Rep method because the B-Rep is basically used for homogeneous models. In addition, meshes are required to analyze a property of a model when the finite elements method is applied. However, the process of generating meshes from B-Rep is cumbersome and sometimes difficult especially when the model is deformed as time goes by because the topology of deforming meshes are changed. To overcome those problems in modeling and analysis including homogeneous and heterogeneous materials, we suggest a unified modeling and analysis method based on implicit representation of the model using R-function which is suggested by Rvachev. For implicit modeling of an object a distance field is approximated and blended for a complex object. Using the implicit function mesh-free analysis is possible where meshes are not necessary. Generally mesh-free analysis requires heavy computational cost compared to a finite elements method. To improve the computing time of function evaluation, we utilize GPU programming. Finally, we give an example of a simple pipe design problem and show modeling and analysis process using our unified modeling and analysis method.

• Honam Mathematical Journal
• /
• v.43 no.2
• /
• pp.183-197
• /
• 2021

Recently several authors have extended the Beta function by using its integral representation. However, in many cases no expression of these extended functions in terms of classic special functions is known. In the present paper, we introduce a further extension by defining a family of functions G_r,s : ℝ^*₊ → ℂ, with r, s ∈ ℂ and ℜ(r) > 0. For given r, s, we prove that this function satisfies a second-order linear differential equation with rational coefficients. Solving this ODE, we express G_r,s as a combination of confluent hypergeometric functions. From this we deduce a new integral relation satisfied by Tricomi's function. We then investigate additional specific properties of G_r,1 which take the form of new non trivial integral relations involving exponential and error functions. We discuss the connection between G_r,1 and Stokes' first problem (or Rayleigh problem) in fluid mechanics which consists in determining the flow created by the movement of an infinitely long plate. For $r{\in}{\frac{1}{2}}{\mathbb{N}}^*$, we find additional relations between G_r,1 and Hermite polynomials. In view of these results, we believe the family of extended beta functions G_r,s will find further applications in two directions: (i) for improving our knowledge of confluent hypergeometric functions and Tricomi's function, (ii) and for engineering and physics problems.

• Structural Engineering and Mechanics
• /
• v.75 no.4
• /
• pp.519-527
• /
• 2020

Dynamic analysis of a vehicle-track coupling system is important to structural design, damage detection and condition assessment of the structural system. Deterministic analysis of the vehicle-track coupling system has been extensively studied in the past, however, the structural parameters of the coupling system have uncertainties in engineering practices. It is essential to treat the parameters of the vehicle-track coupling system with consideration of uncertainties. In this paper, a method for predicting the bounds of the vehicle-track coupling system responses with uncertain parameters is presented. The uncertain system parameters are modeled as fuzzy variables instead of conventional random variables with known probability distributions. Then, the dynamic response functions of the coupling system are transformed into a component function based on the high dimensional representation approximation. The Lagrange interpolation method is used to approximate the component function. Finally, the bounds of the system's dynamic responses can be predicted by using Monte Carlo method for the interpolation polynomials of the Lagrange interpolation function. A numerical example is introduced to illustrate the ability of the proposed method to predict the bounds of the system's dynamic responses, and the results are compared with the direct Monte Carlo method. The results show that the proposed method is effective and efficient to predict the bounds of the system's dynamic responses with fuzzy variables.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2004.04a
• /
• pp.239-243
• /
• 2004

It does not exist almost that Most satellite image has both high spectral and spatial resolution. In order to apply the satellite image for to be actual, we need numerical and analytical technique development to improve the resolution. Specially in the function of solid illustration, we represent the solid image through the image generation to solid screen. The main function includes magnification, reduction, screen center movement, Panning, territory magnification. The method to process the image includes histogram and contrast modulation. Afterwards, we will develop the function includes 3-dimension cursor to control the elevation position and calculate the ground coordination automatically. There is the layer control includes the representation and the edition of 3D vector, extraction the Z value by On the Ground and digital elevation.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.1
• /
• pp.251-268
• /
• 1996

In this paper, a higher-order feedforward neural network called ridge polynomial network (RPN) which shows good approximation capability for nonlnear continuous functions defined on compact subsets in multi-dimensional Euclidean spaces, is presented. This network provides more efficient and regular structure as compared to ordinary higher-order feedforward networks based on Gabor-Kolmogrov polynomial expansions, while maintating their fast learning property. the ridge polynomial network is a generalization of the pi-sigma network (PSN) and uses a specialform of ridge polynomials. It is shown that any multivariate polynomial can be exactly represented in this form, and thus realized by a RPN. The approximation capability of the RPNs for arbitrary continuous functions is shown by this representation theorem and the classical weierstrass polynomial approximation theorem. The RPN provides a natural mechanism for incremental function approximation based on learning algorithm of the PSN. Simulation results on several applications such as multivariate function approximation and pattern classification assert nonlinear approximation capability of the RPN.

• Proceedings of the Korea Water Resources Association Conference
• /
• 1983.07a
• /
• pp.78-79
• /
• 1983

It is well known that small amplitude wave theory, a first approximation to the complete theoretical description of wave behavior, yields a maximum investment in mathematical endeavor. But, if the wave amplitude is large, the small amplitude considerations are not valid, and finite amplitude wave theory which retains higher-order terms to obtain an accurate representation of the wave motion is numercal theory. The Stream function wave theory, one of the numerical methods, was developed by Dean for use with asymmetric measured wave profiles and with symmetric theoretical wave profiles. Dalrymple later improved the comjputational procedure by adding two Lagrangian constraints so that more efficient convergence of the iterative numerical method to a specified wave heigh and to a zero mean free surface displacement resulted. This paper introduces in details the Dean and Darlymple Stream Function Method in case of the symmetric theoretical wave, because in design purposes, wave height and wave period are given.