• Structural Engineering and Mechanics
• /
• v.66 no.1
• /
• pp.15-25
• /
• 2018

Post-construction subgrade settlement especially differential settlement, has become a key issue in construction and operation of non-ballasted track on high-speed railway soil subgrade, which may also affect the dynamic performance of passing trains. To estimate the effect of differential subgrade settlement on the mechanical behaviors of the vehicle-slab track system, a detailed model considering nonlinear subgrade support and initial track state due to track self-weight is developed. Accordingly, analysis aiming at a typical high-speed vehicle coupled with a deteriorated slab track owing to differential subgrade settlement is carried out, in terms of two aspects: (i) determination of an initial mapping relationship between subgrade settlement and track deflections as well as contact state between track and subgrade based on a semi-analytical method; (ii) simulation of dynamic performance of the coupled system by employing a time integration approach. The investigation indicates that subgrade settlement results in additional track irregularity, and locally, the contact between the concrete track and the soil subgrade is prone to failure. Moreover, wheel-rail interaction is significantly exacerbated by the track degradation and abnormal responses occur as a result of the unsupported areas. Distributions of interlaminar contact forces in track system vary dramatically due to the combined effect of track deterioration and dynamic load. These may not only intensify the dynamic responses of the coupled system, but also have impacts on the long-term behavior of the track components.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.8 no.2
• /
• pp.19-25
• /
• 2004

The linear analysis for the balance of linear momentum of a structure is relatively easy to perform, but the error becomes large when the structure experiences large deformation. Therefore, the material and geometric nonlinearity need to be considered for the precise calculations in that case. The plastic flow of a ductile steel-like metal mainly transforms its dissipated mechanical energy into heat, which transfers under the first and second law of thermodynamics. This heat increases the temperature of the material and the strength of the material decreases accordingly, which affects mechanical behavior of the given structure. This paper presents a finite-strain thermo-elasto-plastic steel model. This model can handle large deformation and thermal load simultaneously, which is common during earthquake periods. Two 3-dimensional finite element analyses verify this formulation.

• Korean Journal of Optics and Photonics
• /
• v.14 no.6
• /
• pp.623-629
• /
• 2003

In this paper, we have proposed a full phase encryption scheme based on phase-encoded XOR operation. The proposed scheme encrypts a gray-level image by slicing an original image and combining with XORed images which resulted from phase-encoded XOR operations between sliced images and phase-encoded binary random images. Then we produce an encrypted image by combining only XORed images and a key image by only phase-encoded binary random images. The encrypted image and key image are converted into encrypted data and key data by a phase-encoding method. The merits are that the proposed encryption scheme can basically fulfill a high-level encryption using a full phase encryption scheme which has nonlinear and invisible characteristics. The scheme also improves security by encrypting the phase information before full phase encryption. The decryption system based on the principle of interference between a reference wave and a direct pixel-to-pixel mapping image of encrypted data with key data can be simply implemented using a phase-visualization system. Simulation results indicate that our proposed encryption scheme is effective and simple for a gray-scale image and optical decryption system.

• Journal of Broadcast Engineering
• /
• v.24 no.4
• /
• pp.633-642
• /
• 2019

Recently, the convolution neural network (CNN) model at a single image super-resolution (SISR) have been very successful. The residual learning method can improve training stability and network performance in CNN. In this paper, we propose a SISR using recursive residual network architecture by introducing dense skip connections for learning nonlinear mapping from low-resolution input image to high-resolution target image. The proposed SISR method adopts a method of the recursive residual learning to mitigate the difficulty of the deep network training and remove unnecessary modules for easier to optimize in CNN layers because of the concise and compact recursive network via dense skip connection method. The proposed method not only alleviates the vanishing-gradient problem of a very deep network, but also get the outstanding performance with low complexity of neural network, which allows the neural network to perform training, thereby exhibiting improved performance of SISR method.

• Geomechanics and Engineering
• /
• v.25 no.3
• /
• pp.195-205
• /
• 2021

Time effect on the deformation and strength characteristics of geogrid reinforced sand retaining wall has become an important issue in geotechnical and transportation engineering. Three physical model tests on geogrid reinforced sand retaining walls performed under various loading conditions were simulated to study their rate-dependent behaviors, using the presented nonlinear finite element method (FEM) analysis procedure. This FEM was based on the dynamic relaxation method and return mapping scheme, in which the combined effects of the rate-dependent behaviors of both the backfill soil and the geosynthetic reinforcement have been included. The rate-dependent behaviors of sands and geogrids should be attributed to the viscous property of materials, which can be described by the unified three-component elasto-viscoplastic constitutive model. By comparing the FEM simulations and the test results, it can be found that the present FEM was able to be successfully extended to the boundary value problems of geosynthetic reinforced soil retaining walls. The deformation and strength characteristics of the geogrid reinforced sand retaining walls can be well reproduced. Loading rate effect, the trends of jump in footing pressure upon the step-changes in the loading rate, occurred not only on sands and geogrids but also on geogrid reinforced sands retaining walls. The lateral earth pressure distributions against the back of retaining wall, the local tensile force in the geogrid arranged in the retaining wall and the local stresses beneath the footing under various loading conditions can also be predicted well in the FEM simulations.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.127-146
• /
• 2022

Recently, as word embedding has shown excellent performance in various tasks of deep learning-based natural language processing, researches on the advancement and application of word, sentence, and document embedding are being actively conducted. Among them, cross-language transfer, which enables semantic exchange between different languages, is growing simultaneously with the development of embedding models. Academia's interests in vector alignment are growing with the expectation that it can be applied to various embedding-based analysis. In particular, vector alignment is expected to be applied to mapping between specialized domains and generalized domains. In other words, it is expected that it will be possible to map the vocabulary of specialized fields such as R&D, medicine, and law into the space of the pre-trained language model learned with huge volume of general-purpose documents, or provide a clue for mapping vocabulary between mutually different specialized fields. However, since linear-based vector alignment which has been mainly studied in academia basically assumes statistical linearity, it tends to simplify the vector space. This essentially assumes that different types of vector spaces are geometrically similar, which yields a limitation that it causes inevitable distortion in the alignment process. To overcome this limitation, we propose a deep learning-based vector alignment methodology that effectively learns the nonlinearity of data. The proposed methodology consists of sequential learning of a skip-connected autoencoder and a regression model to align the specialized word embedding expressed in each space to the general embedding space. Finally, through the inference of the two trained models, the specialized vocabulary can be aligned in the general space. To verify the performance of the proposed methodology, an experiment was performed on a total of 77,578 documents in the field of 'health care' among national R&D tasks performed from 2011 to 2020. As a result, it was confirmed that the proposed methodology showed superior performance in terms of cosine similarity compared to the existing linear vector alignment.

• The Korean Journal of Nuclear Medicine
• /
• v.34 no.1
• /
• pp.10-21
• /
• 2000

Purpose: Episodic memory is described as an 'autobiographical' memory responsible for storing a record of the events in our lives. We performed functional brain activation study using ${H_2}^{15}O$ PET to reveal the neural basis of the encoding and the retrieval of episodic memory in human normal volunteers. Materials and Methods: Four repeated ${H_2}^{15}O$ PET scans with two reference and two activation tasks were performed on 6 normal volunteers to activate brain areas engaged in encoding and retrieval with verbal materials. Images from the same subject were spatially registered and normalized using linear and nonlinear transformation. Using the means and variances for every condition which were adjusted with analysis of covariance, t-statistic analysis were performed voxel-wise. Results: Encoding of episodic memory activated the opercular and triangular parts of left inferior frontal gyrus, right prefrontal cortex, medial frontal area, cingulate gyrus, posterior middle and inferior temporal gyri, and cerebellum, and both primary visual and visual association areas. Retrieval of episodic memory activated the triangular part of left inferior frontal gyrus and inferior temporal gyrus, right prefrontal cortex and medial temporal area, and both cerebellum and primary visual and visual association areas. The activations in the opercular part of left inferior frontal gyrus and the right prefrontal cortex meant the essential role of these areas in the encoding and retrieval of episodic memory. Conclusion: We could localize the neural basis of the encoding and retrieval of episodic memory using ${H_2}^{15}O$ PET, which was partly consistent with the hypothesis of hemispheric encoding/retrieval asymmetry.

• Nuclear Medicine and Molecular Imaging
• /
• v.42 no.2
• /
• pp.153-163
• /
• 2008

Medical imaging modalities to image either anatomical structure or functional processes have developed along somewhat independent paths. Functional images with single photon emission computed tomography (SPECT) and positron emission tomography (PET) are playing an increasingly important role in the diagnosis and staging of malignant disease, image-guided therapy planning, and treatment monitoring. SPECT and PET complement the more conventional anatomic imaging modalities of computed tomography (CT) and magnetic resonance (MR) imaging. When the functional imaging modality was combined with the anatomic imaging modality, the multimodality can help both identify and localize functional abnormalities. Combining PET with a high-resolution anatomical imaging modality such as CT can resolve the localization issue as long as the images from the two modalities are accurately coregistered. Software-based registration techniques have difficulty accounting for differences in patient positioning and involuntary movement of internal organs, often necessitating labor-intensive nonlinear mapping that may not converge to a satisfactory result. These challenges have recently been addressed by the introduction of the combined PET/CT scanner and SPECT/CT scanner, a hardware-oriented approach to image fusion. Combined PET/CT and SPECT/CT devices are playing an increasingly important role in the diagnosis and staging of human disease. The paper will review the development of multi modality instrumentations for clinical use from conception to present-day technology and the application software.

• Journal of Internet Computing and Services
• /
• v.19 no.5
• /
• pp.43-54
• /
• 2018

Images taken in haze weather are characteristic of low contrast and poor visibility. The process of reconstructing clear-weather image from a hazy image is called dehazing. The main challenge of image dehazing is to estimate the transmission map or depth map for an input hazy image. In this paper, we propose a single image dehazing method by utilizing the Generative Adversarial Network(GAN) for accurate depth map estimation. The proposed GAN model is trained to learn a nonlinear mapping between the input hazy image and corresponding depth map. With the trained model, first the depth map of the input hazy image is estimated and used to compute the transmission map. Then a guided filter is utilized to preserve the important edge information of the hazy image, thus obtaining a refined transmission map. Finally, the haze-free image is recovered via atmospheric scattering model. Although the proposed GAN model is trained on synthetic indoor images, it can be applied to real hazy images. The experimental results demonstrate that the proposed method achieves superior dehazing results against the state-of-the-art algorithms on both the real hazy images and the synthetic hazy images, in terms of quantitative performance and visual performance.

• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.378-390
• /
• 2022

With the recent development of deep composite multiplication neural network learning, deep learning techniques applied to single-image super-resolution have shown good results, and the strong expression ability of deep networks has enabled complex nonlinear mapping between low-resolution and high-resolution images. However, there are limitations in applying it to real-time or low-power devices with increasing parameters and computational amounts due to excessive use of composite multiplication neural networks. This paper uses blocks that extract hierarchical characteristics little by little using information distillation and suggests the Recursive Distillation Super Resolution Network (RDSRN), a lightweight network that improves performance by making more accurate high frequency components through high frequency residual purification blocks. It was confirmed that the proposed network restores images of similar quality compared to RDN, restores images 3.5 times faster with about 32 times fewer parameters and about 10 times less computation, and produces 0.16 dB better performance with about 2.2 times less parameters and 1.8 times faster processing time than the existing lightweight network CARN.