• Communications for Statistical Applications and Methods
• /
• v.30 no.6
• /
• pp.589-603
• /
• 2023

Deep generative models target to infer the underlying true data distribution, and it leads to a huge success in generating fake-but-realistic data. Regarding such a perspective, the data attributes can be a crucial factor in the data generation process since non-existent counterfactual samples can be generated by altering certain factors. For example, we can generate new portrait images by flipping the gender attribute or altering the hair color attributes. This paper proposes counterfactual disentangled variational autoencoder generative adversarial networks (CDVAE-GAN), specialized for data attribute level counterfactual data generation. The structure of the proposed CDVAE-GAN consists of variational autoencoders and generative adversarial networks. Specifically, we adopt a Gaussian variational autoencoder to extract low-dimensional disentangled data features and auxiliary Bernoulli latent variables to model the data attributes separately. Also, we utilize a generative adversarial network to generate data with high fidelity. By enjoying the benefits of the variational autoencoder with the additional Bernoulli latent variables and the generative adversarial network, the proposed CDVAE-GAN can control the data attributes, and it enables producing counterfactual data. Our experimental result on the CelebA dataset qualitatively shows that the generated samples from CDVAE-GAN are realistic. Also, the quantitative results support that the proposed model can produce data that can deceive other machine learning classifiers with the altered data attributes.

• Journal of the Korean Society of Radiology
• /
• v.10 no.4
• /
• pp.263-270
• /
• 2016

This paper suggests techniques to enhance coding time which is a problem in traditional fractal compression and to improve fidelity of reconstructed images by determining fractal coefficient through adaptive selection of block approximation formula. First, to reduce coding time, we construct a linear list of domain blocks of which characteristics is given by their luminance and variance and then we control block searching time according to the first permissible threshold value. Next, when employing three-level block partition, if a range block of minimum partition level cannot find a domain block which has a satisfying approximation error, There applied to 24-bpp color image compression and image techniques. The result did not occur a loss in the image quality of the image when using the encoding method, such as almost to the color in the YIQ image compression rate and image quality, such as RGB images and showed good.

• Journal of Animal Reproduction and Biotechnology
• /
• v.36 no.3
• /
• pp.121-128
• /
• 2021

Increasing the efficiency of HR (homologous recombination) is important for a successful knock-in. Rad51 is mainly involved in homologous recombination and is associated with strand invasion. The HR-related mismatch repair system maintains HR fidelity by heteroduplex rejection and repair. Therefore, the purpose of this study is to control Rad51, which plays a critical role in HR, through UV-induced DNA damage. It is also to confirm the effect on the expression of MMR related genes (Msh2, Msh3, Msh6, Mlh1, Pms2) and HR-related genes closely related to HR through treatment with the MMR inhibitor CdCl₂. The mRNA expression of Rad51 gene was confirmed in both HC11 cells and mouse testes, but the mRNA expression of Dmc1 gene was confirmed only in mouse testes. The protein expression of Rad51 and Dmc1 gene increased in UV-irradiated HC11 cells. After 72 hours of treatment with 1 ㎛ of CdCl₂, the mRNA expression level of Msh3, Pms2, and Rad51 decreased, but the mRNA expression level of Msh6 and Mlh1 increased in HC11 cells. There was no significant difference in Msh2 mRNA expression between CdCl₂ untreated-group and the 72 hours treated group. In conclusion, HR-related gene (Rad51) was increased by UV-induced DNA damage. Treatment of the MMR inhibitor CdCl₂ in HC11 cells decreased the mRNA expression of Rad51.

• Journal of the Korea Society for Simulation
• /
• v.28 no.2
• /
• pp.89-95
• /
• 2019

This paper introduces the simulation concepts and technical approach of underwater weapon system performance analysis simulator, especially focused on probabilistic target detection concepts. We calculated the signal excess (SE) value using SONAR equation, then derived the probability density function(PDF) for target presence($H_1$) or absence($H_0$) cases, respectively. With the Neyman-Pearson detector criterion, we got the probability of detection($P_D$) while satisfying the given probability of false alarm($P_{FA}$). At every instance of simulation, target detection is decided in the probabilistic perspective. With the proposed detection implementation, we improved the model fidelity so that it could support the tactical decision during the operation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.1
• /
• pp.192-210
• /
• 2024

Most frequency-domain remote sensing image watermarking algorithms embed watermarks at random locations, which have negative impact on the watermark invisibility. In this study, we propose an adaptive watermarking scheme for remote sensing images that considers the information complexity to select where to embed watermarks to improve watermark invisibility without affecting algorithm robustness. The scheme converts remote sensing images from RGB to YCbCr color space, performs two-level DWT on luminance Y, and selects the high frequency coefficient of the low frequency component (HH^Y₂) as the watermark embedding domain. To achieve adaptive embedding, HH^Y₂ is divided into several 8*8 blocks, the entropy of each sub-block is calculated, and the block with the maximum entropy is chosen as the watermark embedding location. During embedding phase, the watermark image is also decomposed by two-level DWT, and the resulting high frequency coefficient (HH^W₂) is then embedded into the block with maximum entropy using α- blending. The experimental results show that the watermarked remote sensing images have high fidelity, indicating good invisibility. Under varying degrees of geometric, cropping, filtering, and noise attacks, the proposed watermarking can always extract high identifiable watermark images. Moreover, it is extremely stable and impervious to attack intensity interference.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.41 no.3
• /
• pp.173-184
• /
• 2013

A multi-level design optimization framework for aerodynamic design of rotary wing such as propeller and helicopter rotor blades is presented in this study. Strategy of the proposed framework is to enhance aerodynamic performance by sequentially applying the planform and sectional design optimization. In the first level of a planform design, we used a genetic algorithm and blade element momentum theory (BEMT) based on two-dimensional aerodynamic database to find optimal planform variables. After an initial planform design, local flow conditions of blade sections are analyzed using high-fidelity CFD methods. During the next level, a sectional design optimization is conducted using two dimensional Navier-Stokes analysis and a gradient based optimization algorithm. When optimal airfoil shape is determined at the several spanwise locations, a planform design is performed again. Through this iterative design process, not only an optimal flow condition but also an optimal shape of an EAV propeller blade is obtained. To validate the optimized propeller-blade design, it is tested in wind-tunnel facility with different flow conditions. An efficiency, which is slightly less than the expected improvement of 7% predicted by our proposed design framework but is still satisfactory to enhance the aerodynamic performance of EAV system.

• The Journal of the Convergence on Culture Technology
• /
• v.6 no.4
• /
• pp.79-84
• /
• 2020

This study aims to establish the lifelong education for the disabled by exploring the academic nature and identity of lifelong education for the disabled from the perspective of convergence, and to guard against a convergence academic system at the level of imitation based on the meaning and function of the established lifelong education for the disabled. The research method consisted of a procedure in parallel with expert meetings based on literature analysis. The content of this study primarily explored the universality and specificity of lifelong education for persons with disabilities in the dimension that it should have an academic essence and identity as a category in which lifelong education for persons with disabilities is fused with several adjacent fields such as special education, centering on the perspective tailored to disability. In addition, the result of exploring the structure system in which the meaning through the universality and specificity of lifelong education for the disabled can be combined not only at the academic level based on theory but also at the functional level based on practice is also presented as the content of the study. As a result of the study, it was analyzed that lifelong education for the disabled can build universality and specificity in the dimension of convergent inquiry, and can establish a support system at a comprehensive level of practice according to the fidelity of the above.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.11
• /
• pp.1021-1030
• /
• 2010

In the sensor network, many tiny nodes construct Ad-Hoc network using wireless interface. As this type of system consists of thousands of nodes, managing each sensor node in real world after deploying them is very difficult. In order to install the sensor network successfully, it is necessary to verify its software using a simulator beforehand. In fact Sensor network simulators require high fidelity and timing accuracy to be used as a design, implementation, and evaluation tool of wireless sensor networks. Cycle-accurate, instruction-level simulation is the known solution for those purposes. In this paper, we developed an instruction-level sensor network simulator for Telos sensor node as named TeloSlM. It consists of MSP430 and CC2420. Recently, Telos is the most popular mote because MSP430 can consume the minimum energy in recent motes and CC2420 can support Zigbee. So that TeloSlM can provide the easy way for the developers to verify software. It is cycle-accurate in instruction-level simulator that is indispensable for OS and the specific functions and can simulate scalable sensor network at the same time. In addition, TeloSlM provides the GUI Tool to show result easily.

• Journal of computational fluids engineering
• /
• v.20 no.2
• /
• pp.7-15
• /
• 2015

This paper attempts to evaluate the accuracy and efficiency of a design optimization procedure by combining wavelets-based multi resolution analysis method and proper orthogonal decomposition (POD) technique. Aerodynamic design procedure calls for high fidelity computational fluid dynamic (CFD) simulations and the consideration of large number of flow conditions and design constraints. Thus, even with significant computing power advancement, current level of integrated design process requires substantial computing time and resources. POD reduces the degree of freedom of full system by conducting singular value decomposition for various field simulations. In this research, POD combined Design Optimization model is proposed and its efficiency and accuracy are to be evaluated. For additional efficiency improvement of the procedure, multi resolution analysis method is also being employed during snapshot constructions (POD training period). The proposed design procedure was applied to the optimization of wing aerodynamic performance. Throughout the research, it was confirmed that the POD/MRA design procedure could significantly reduce the total design turnaround time and also capture all detailed complex flow features as in full order analysis.

• Smart Structures and Systems
• /
• v.25 no.2
• /
• pp.155-167
• /
• 2020

This paper demonstrates a real-time hybrid substructuring (RTHS) shake table test to evaluate the seismic performance of a base isolated building. Since RTHS involves a feedback loop in the test implementation, the frequency dependent magnitude and inherent time delay of the actuator dynamics can introduce inaccuracy and instability. The paper presents a robust stability and performance analysis method for the RTHS test. The robust stability method involves casting the actuator dynamics as a multiplicative uncertainty and applying the small gain theorem to derive the sufficient conditions for robust stability and performance. The attractive feature of this robust stability and performance analysis method is that it accommodates linearized modeled or measured frequency response functions for both the physical substructure and actuator dynamics. Significant experimental research has been conducted on base isolators and dampers toward developing high fidelity numerical models. Shake table testing, where the building superstructure is tested while the isolation layer is numerically modeled, can allow for a range of isolation strategies to be examined for a single shake table experiment. Further, recent concerns in base isolation for long period, long duration earthquakes necessitate adding damping at the isolation layer, which can allow higher frequency energy to be transmitted into the superstructure and can result in damage to structural and nonstructural components that can be difficult to numerically model and accurately predict. As such, physical testing of the superstructure while numerically modeling the isolation layer may be desired. The RTHS approach has been previously proposed for base isolated buildings, however, to date it has not been conducted on a base isolated structure isolated at the ground level and where the isolation layer itself is numerically simulated. This configuration provides multiple challenges in the RTHS stability associated with higher physical substructure frequencies and a low numerical to physical mass ratio. This paper demonstrates a base isolated RTHS test and the robust stability and performance analysis necessary to ensure the stability and accuracy. The tests consist of a scaled idealized 4-story superstructure building model placed directly onto a shake table and the isolation layer simulated in MATLAB/Simulink using a dSpace real-time controller.