• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.125-128
• /
• 1999

근거리음장 측정으로부터 원거리음장을 예측할 수 있는 이론적 모델 및 실험적 연구가 수행되고 있다. 본 논문에서 채택한 방법은 Helmholtz 적분식의 간단한 모델로써 단일주파수 코히어런트 음원의 근거리측정음장으로부터 원거리음장 예측이 가능한 기술이다. 이러한 단일 주파수, 코히어런트 음원의 원거리 수평방향지향성을 예측하기 위한 모델식 및 무향실에서 수행한 검증실험 결과를 소개한다 제안된 방법은 일반 소음원은 물론 잠수함 등의 수평방향 지향성의 예측에 적용가능하다고 판단된다.

• Journal of KIISE
• /
• v.42 no.6
• /
• pp.817-822
• /
• 2015

Relation extraction is an important information extraction technique that can be widely used in areas such as question-answering and knowledge population. Previous studies on relation extraction have been based on supervised machine learning models that need a large amount of training data manually annotated with relation categories. Recently, to reduce the manual annotation efforts for constructing training data, distant supervision methods have been proposed. However, these methods suffer from a drawback: it is difficult to use these methods for collecting negative training data that are necessary for resolving classification problems. To overcome this drawback, we propose a one-class classification model that can be trained without using negative data. The proposed model determines whether an input data item is included in an inner category by using a similarity measure based on lexical information and syntactic patterns in a vector space. In the experiments conducted in this study, the proposed model showed higher performance (an F1-score of 0.6509 and an accuracy of 0.6833) than a representative one-class classification model, one-class SVM(Support Vector Machine).

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.1
• /
• pp.40-46
• /
• 2004

The equivalent source method (ESM) uses two groups of equivalent source positions. One group includes the first order images of the sound source inside the enclosure. The positions of the other group are usually on a spherical surface some distance outside the enclosure. A proper selection of the positions for the far field sources could greatly improve the performance of the modeling accuracy and reduce the number of the sources to achieve the required accuracy. This study uses optimally distributed far field source positions on the surface of enlarged version of the rectangular enclosure instead of using typical spherical distribution. Simulations using various sizes of the box shaped distribution are executed and optimal positions are searched using an optimization technique from the larger number of candidate positions. The results of using these far field source positions are compared and analyzed.

• Journal of Radiation Protection and Research
• /
• v.23 no.1
• /
• pp.1-6
• /
• 1998

A diffusion model of radioactive liquid effluents is developed and applied for YGN NPP's site, based on the Gaussian plume type model. Due to the complexity of oceanic diffusion characteristics of YGN site, a simple and reliable statistical model based on Reg. Guide 1.113 is developed. Also, a computer code package to calculate dilution factors as a function of plant operation conditions and pathway of radioactive materials. A liquid effluents diffusion model is developed by dividing the diffusion range into two categories, i. e, a near field mixing region and a far field mixing region. In the near field, the initial mixing is affected by a buoyance force, a high initial turbulence and momentum which is characterized by a plant operation condition and environmental conditions. The far field mixing is similar to gaseous effluents diffusion. So, beyond the near field region, wellknown Gaussian plume model was adopted. A different area averages of Gaussian plume equation was taken for each radioactive exposure pathway. As a result, we can get different dilution factors for different pathways. Results shows that present dilution factors used for YGN ODCM is too much overestimated compared with dilution factors calculated with the developed model.

• Explosives and Blasting
• /
• v.29 no.2
• /
• pp.32-42
• /
• 2011

For blasting vibration analysis, there are mostly two methods, firstly, far field vibration analysis to estimate limit of building damage, secondly, near field subterranean vibration analysis to estimate rock damage and subterranean structure facilities. The former has been mainly used in our country and aboard, the latter is mostly analysed by using Homberg's model. In this model, basic input data are acquired in far field surface vibration. But in the consideration of the results of being conducted in this area over the past few decade, it is required that Homberg's model is modified. For the purpose of this, measurements of near field vibration were first conducted in our country. But it was only proposed the measurement method and the method of analysis or prediction was not suggested. Accordingly, in this paper, measurements of near field subterranean vibration were conducted and the method of analysis or prediction of near field subterranean vibration would be suggested.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.1019-1024
• /
• 2006

각종 온라인 문화가 발달하는 가운데 원거리에서 의사 소통을 가능하게 하는 도구에 관한 관심도 커지게 되었다. 이러한 도구 중에서 2 차원 그림 저작을 지원하는 도구에 기인하여, 본 논문에서는 3 차원 모델을 만들어 내고 조작함으로써 원거리의 사용자와 의사 소통할 수 있는 시스템을 제안한다. 실시간에 수행되는 시스템의 특성상, 3 차원 모델링은 간편하고 직관적으로 수행될 수 있도록 하며, 병행 수행 제어를 고려하여 구현한다. 본 시스템에서의 모델링은 주어진 primitive를 이용하거나 사전에 제작된 mesh를 불러오는 것으로 수행할 수 있으며, 사용자가 직접 그려주는 간략한 외곽선 스케치를 기반으로 하여 3 차원 모델을 생성할 수도 있다. 본 시스템에서 모델의 생성, 제어 등의 모든 연산은 다양한 병행 수행 제어 전략을 이용하여 실시간에 수행된다.

• The Journal of the Acoustical Society of Korea
• /
• v.34 no.6
• /
• pp.455-462
• /
• 2015

It is well known that warm eddy is frequently developed through the year in the East Sea. The warm eddy may affect sound propagation due to changes of sound velocity structures in the sea water. To verify the effects of the warm eddy for long-range sound propagation, transmission loss and performance surface, which were used mean direct signal excess range generated by sound propagation modeling using re-analyzed climatology data on March 23th in 2007 were analysed. From these analyses, we found that characteristics of sound propagation in the sea water are changed by the warm eddy, and boundaries of the warm eddy act as a barrier for long-range sound propagation. Furthermore, these disadvantages of the eddy related to sound propagation were increased when the sea bottom depth is shallow.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.11
• /
• pp.2862-2868
• /
• 1996

A sueueing model is developed and analyzed to evaluate the performance of a local and a remote destination of one multicast connection subject to a simple two-level(binary) feedback-based adaptive rate control scheme. The reate control scheme is a simple extension of the one for point-to-point case. Individual cell loss probabilities at the local and remote destinations are obtained to study the fairness of the adaptive rate control scheme. Numerical results show that this simple adaptive rate control scheme, which is fair when it is applied to a multicast connection consisting of destinations with equal propagation delays, bears bias against remote destination.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.2
• /
• pp.125-133
• /
• 2004

Most of the way shape estimation method using reference sources assume that the reference sources are in the farfield. That is, the reference sources are assumed to be far from the array. However, in applications of the array with reference sources, the reference sources are not far from the way, so that in practical ocean environments, the conventional method using farfield source model fail to estimate the positions of the hydrophones. In this paper, based on the nearfield source model, a subspace-based array shape estimation method was proposed. In the proposed method, nearfield reference source is modeled using the differential time delay at each hydrophone, and nearfield parameters are derived. Using these parameters, a subspace-based array shape estimation method that generalizes the existing farfield subspace fitting method which can work regardless of the range of the source is proposed. The Cramer-Rao lower bound for the proposed method is investigated. The results of the numerical experiments indicate that the proposed method performs well in estimating the shape of a perturbed way regardless of the ranges of the reference sources.

• Journal of Broadcast Engineering
• /
• v.28 no.3
• /
• pp.329-332
• /
• 2023

3D Object Detection using deep neural network has been developed a lot for obstacle detection in autonomous vehicles because it can recognize not only the class of target object but also the distance from the object. But in the case of 3D Object Detection models, the detection performance for distant objects is lower than that for nearby objects, which is a critical issue for autonomous vehicles. In this paper, we introduce a technique that increases the performance of 3D object detection models, particularly in recognizing distant objects, by generating virtual 3D vehicle data and adding it to the dataset used for model training. We used a spherical point tracing method that leverages the characteristics of 3D LiDAR sensor data to create virtual vehicles that closely resemble real ones, and we demonstrated the validity of the virtual data by using it to improve recognition performance for objects at all distances in model training.