• Journal of the Korea Convergence Society
• /
• 제12권11호
• /
• pp.45-51
• /
• 2021

With the development of deep learning, semantic segmentation methods are being studied in various fields. There is a problem that segmenation accuracy drops in fields that require accuracy such as medical image analysis. In this paper, we improved PSPNet, which is a deep learning based segmentation method to minimized the loss of features during semantic segmentation. Conventional deep learning based segmentation methods result in lower resolution and loss of object features during feature extraction and compression. Due to these losses, the edge and the internal information of the object are lost, and there is a problem that the accuracy at the time of object segmentation is lowered. To solve these problems, we improved PSPNet, which is a semantic segmentation model. The multi-scale attention proposed to the conventional PSPNet was added to prevent feature loss of objects. The feature purification process was performed by applying the attention method to the conventional PPM module. By suppressing unnecessary feature information, eadg and texture information was improved. The proposed method trained on the Cityscapes dataset and use the segmentation index MIoU for quantitative evaluation. As a result of the experiment, the segmentation accuracy was improved by about 1.5% compared to the conventional PSPNet.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권10호
• /
• pp.4976-4994
• /
• 2018

The diffusion-based salient object detection methods have shown excellent detection results and more efficient computation in recent years. However, the current diffusion-based salient object detection methods still have disadvantage of detecting the object appearing at the image boundaries and different scales. To address the above mentioned issues, this paper proposes a multi-scale diffusion-based salient object detection algorithm with background and objectness seeds. In specific, the image is firstly over-segmented at several scales. Secondly, the background and objectness saliency of each superpixel is then calculated and fused in each scale. Thirdly, manifold ranking method is chosen to propagate the Bayessian fusion of background and objectness saliency to the whole image. Finally, the pixel-level saliency map is constructed by weighted summation of saliency values under different scales. We evaluate our salient object detection algorithm with other 24 state-of-the-art methods on four public benchmark datasets, i.e., ASD, SED1, SED2 and SOD. The results show that the proposed method performs favorably against 24 state-of-the-art salient object detection approaches in term of popular measures of PR curve and F-measure. And the visual comparison results also show that our method highlights the salient objects more effectively.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 한국전산구조공학회 2007년도 정기 학술대회 논문집
• /
• pp.427-432
• /
• 2007

In general, the response of bulk material is independent of its size when it comes to considering classical elasticity theory. Because the surface to bulk ratio of the large solids is very small, the influence of surface can be negligible. But the surface effect plays important role as the surface to bulk ratio becomes larger, that is, the contribution of the surface effect must be considered in nano-size elements such as thin film or beam structure. Molecular dynamics computation has been a conventional way to analyze these ultra-thin structures but this method is limited to simulate on the order of $10^6-10^8$ atoms for a few nanoseconds, and besides, very time consuming. Analysis of structures in submicro to micro range(thin-film, wire etc.) is difficult with classical molecular dynamics due to the restriction of computing resources and time. Therefore, in this paper, the continuum-based method is considered to simulate the overall physical and mechanical properties of the structures in nano-scale, especially, for the thin-film.

• Journal of computational fluids engineering
• /
• 제2권1호
• /
• pp.21-28
• /
• 1997

This paper presents a two-dimensional horizontal implicit model to general circulation in estuaries and coastal seas. The model is developed in non-orthogonal curvilinear coordinates system, using the Interpolating Matrix Method (IMM), in combination with a technique of multi-domain. In the propose model, the Saint-Venant equations are solved by a splitting-up technique, in the successive steps; convection, diffusion and wave propagation. The ability of the proposed model to deal with full scale nature is illustrated by the interpretation of a dye-tracing experiment in the Gironde estuary.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권6호
• /
• pp.1620-1634
• /
• 2023

This paper studies the task of embedding nodes with multiple graphs representing multiple information channels, which is useful in a large volume of network clustering tasks. By learning a node using multiple graphs, various characteristics of the node can be represented and embedded stably. Existing studies using multi-channel networks have been conducted by integrating heterogeneous graphs or limiting common nodes appearing in multiple graphs to have similar embeddings. Although these methods effectively represent nodes, it also has limitations by assuming that all networks provide the same amount of information. This paper proposes a method to overcome these limitations; The proposed method gives different weights according to the source graph when embedding nodes; the characteristics of the graph with more important information can be reflected more in the node. To this end, a novel method incorporating a multi-channel gate layer is proposed to weigh more important channels and ignore unnecessary data to embed a node with multiple graphs. Empirical experiments demonstrate the effectiveness of the proposed multi-channel-based embedding methods.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• 제45권1호
• /
• pp.25-36
• /
• 2008

Eye localization means localization of the center of the pupils, and is necessary for face recognition and related applications. Most of eye localization methods reported so far still need to be improved about robustness as well as precision for successful applications. In this paper, we propose a robust eye localization method using multi-scale Gabor feature vectors without big computational burden. The eye localization method using Gabor feature vectors is already employed in fuck as EBGM, but the method employed in EBGM is known not to be robust with respect to initial values, illumination, and pose, and may need extensive search range for achieving the required performance, which may cause big computational burden. The proposed method utilizes multi-scale approach. The proposed method first tries to localize eyes in the lower resolution face image by utilizing Gabor Jet similarity between Gabor feature vector at an estimated initial eye coordinates and the Gabor feature vectors in the eye model of the corresponding scale. Then the method localizes eyes in the next scale resolution face image in the same way but with initial eye points estimated from the eye coordinates localized in the lower resolution images. After repeating this process in the same way recursively, the proposed method funally localizes eyes in the original resolution face image. Also, the proposed method provides an effective illumination normalization to make the proposed multi-scale approach more robust to illumination, and additionally applies the illumination normalization technique in the preprocessing stage of the multi-scale approach so that the proposed method enhances the eye detection success rate. Experiment results verify that the proposed eye localization method improves the precision rate without causing big computational overhead compared to other eye localization methods reported in the previous researches and is robust to the variation of post: and illumination.

• The KSFM Journal of Fluid Machinery
• /
• 제17권4호
• /
• pp.19-29
• /
• 2014

In addition to the development of micro hydro turbine, the challenge in micro hydro turbine design as sustainable hydro devices is focused on the optimization of turbine runner blade which have decisive effect on the turbine performance to reach higher efficiency. A multi-objective optimization method to optimize the performance of runner blade of propeller turbine for micro turbine has been studied. For the initial design of planar blade cascade, singularity distribution method and the combination of the Bezier curve parametric technology is used. A non-dominated sorting genetic algorithm II(NSGA II) is developed based on the multi-objective optimization design method. The comparision with model test show that the blade charachteristics is optimized by NSGA-II has a good efficiency and load distribution. From model test and scale up calculation, the maximum prototype efficiency of the runner blade reaches as high as 90.87%.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.690-693
• /
• 2004

This paper proposes a multi-resolution electrostatic potential field (MREPF) based solution to the mobile robot path planning and collision avoidance problem in 2D dynamic environment. The MREPF is an environment method in calculation time and updating field map. The large scale resolution map is added to EPF and this resolution map interacts with the small scale resolution map to find an optimal solution in real time. This approach can be interpreted with Atlantis model. The simulation studies show the efficiency of the proposed algorithm.

• Journal of the Korean Statistical Society
• /
• 제36권2호
• /
• pp.209-228
• /
• 2007

Given scattered surface air temperatures observed by a network of weather stations, it is an important problem to estimate the entire temperature field for every location on the globe. Recently, a multi-scale spherical basis function (SBF) representation was proposed by Li (1999) for representing scattered data on the sphere. However, for a successful application of Li (1999)'s method, some practical issues such as network design, bandwidth selection of SBFs and initial coefficients are to be resolved. This paper proposes automatic procedures to design network and to select bandwidths. This paper also considers a preprocessing problem to obtain a stable initial coefficients from scattered data. Experiments with real temperature data demonstrate the promising empirical properties of the proposed approaches.

• Journal of the Korean Society for Railway
• /
• 제4권2호
• /
• pp.47-54
• /
• 2001

Railway companies have been faced with many changes in the railway environment. To cope immediately with the influence of environment and to improve productivity, an efficient train operation system and related core technologies must be introduced. The railway system is composed of large scale infrastructures and high-cost trains. Simulation method is one of core technologies and also efficient tool for planning and analyzing these kinds of complex system. In this research, we review basic simulation programming models and present a modeling for the elements of railway system such as rail-line infrastructure, train, time table and operational route. Additionally, some considerations on the development of multi-train traffic simulator for KyongBu-line are discussed.