• The Journal of the Acoustical Society of Korea
• /
• v.42 no.4
• /
• pp.357-363
• /
• 2023

In underwater signal processing, separating individual signals from mixed signals has long been a challenge due to low signal quality. The common method using Short-time Fourier transform for spectrogram analysis has faced criticism for its complex parameter optimization and loss of phase data. We propose a Triple-path Recurrent Neural Network, based on the Dual-path Recurrent Neural Network's success in long time series signal processing, to handle three-dimensional tensors from multi-channel sensor input signals. By dividing input signals into short chunks and creating a 3D tensor, the method accounts for relationships within and between chunks and channels, enabling local and global feature learning. The proposed technique demonstrates improved Root Mean Square Error and Scale Invariant Signal to Noise Ratio compared to the existing method.

• Journal of the Chungcheong Mathematical Society
• /
• v.37 no.1
• /
• pp.27-40
• /
• 2024

In this paper, we consider the higher-order HartreeFock equations. The higher-order linear Schrödinger equation was introduced in [5] as the formal finite Taylor expansion of the pseudorelativistic linear Schrödinger equation. In [13], the authors established global-in-time Strichartz estimates for the linear higher-order equations which hold uniformly in the speed of light c ≥ 1 and as their applications they proved the convergence of higher-order Hartree-Fock equations to the corresponding pseudo-relativistic equation on arbitrary time interval as c goes to infinity when the Taylor expansion order is odd. To achieve this, they not only showed the existence of solutions in L² space but also proved that the solutions stay bounded uniformly in c. We address the remaining question on the convergence of higherorder Hartree-Fock equations when the Taylor expansion order is even. The distinguished feature from the odd case is that the group velocity of phase function would be vanishing when the size of frequency is comparable to c. Owing to this property, the kinetic energy of solutions is not coercive and only weaker Strichartz estimates compared to the odd case were obtained in [13]. Thus, we only manage to establish the existence of local solutions in H^s space for s > $\frac{1}{3}$ on a finite time interval [-T, T], however, the time interval does not depend on c and the solutions are bounded uniformly in c. In addition, we provide the convergence result of higher-order Hartree-Fock equations to the pseudo-relativistic equation with the same convergence rate as the odd case, which holds on [-T, T].

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.4
• /
• pp.144-159
• /
• 2014

Techniques for making a single panoramic image using multiple pictures are widely studied in many areas such as computer vision, computer graphics, etc. The panorama image can be applied to various fields like virtual reality, robot vision areas which require wide-angled shots as an useful way to overcome the limitations such as picture-angle, resolutions, and internal informations of an image taken from a single camera. It is so much meaningful in a point that a panoramic image usually provides better immersion feeling than a plain image. Although there are many ways to build a panoramic image, most of them are using the way of extracting feature points and matching points of each images for making a single panoramic image. In addition, those methods use the RANSAC(RANdom SAmple Consensus) algorithm with matching points and the Homography matrix to transform the image. The SURF(Speeded Up Robust Features) algorithm which is used in this paper to extract featuring points uses an image's black and white informations and local spatial informations. The SURF is widely being used since it is very much robust at detecting image's size, view-point changes, and additionally, faster than the SIFT(Scale Invariant Features Transform) algorithm. The SURF has a shortcoming of making an error which results in decreasing the RANSAC algorithm's performance speed when extracting image's feature points. As a result, this may increase the CPU usage occupation rate. The error of detecting matching points may role as a critical reason for disqualifying panoramic image's accuracy and lucidity. In this paper, in order to minimize errors of extracting matching points, we used $3{\times}3$ region's RGB pixel values around the matching points' coordinates to perform intermediate filtering process for removing wrong matching points. We have also presented analysis and evaluation results relating to enhanced working speed for producing a panorama image, CPU usage rate, extracted matching points' decreasing rate and accuracy.

• Journal of Astronomy and Space Sciences
• /
• v.28 no.4
• /
• pp.305-310
• /
• 2011

The total electron content (TEC) using global positioning system (GPS) is analyzed to see the characteristics of ionosphere over King Sejong station (KSJ, geographic latitude $62^{\circ}13'S$, longitude $58^{\circ}47'W$, corrected geomagnetic latitude $48^{\circ}S$) in Antarctic. The GPS operational ratio during the observational period between 2005 and 2009 is 90.1%. The annual variation of the daily mean TEC decreases from January 2005 to February 2009, but increase from the June 2009. In summer (December-February), the seasonal mean TEC values have the maximum of 26.2 ${\pm}$ 2.4 TEC unit (TECU) in 2005 and the minimum of 16.5 ${\pm}$ 2.8 TECU in 2009, and the annual differences decrease from 3.0 TECU (2005-2006) to 1.4 TECU (2008-2009). However, on November 2010, it significantly increases to 22.3 ${\pm}$ 2.8 TECU which is up to 5.8 TECU compared with 2009 in summer. In winter (June-August), the seasonal mean TEC slightly decreases from 13.7 ${\pm}$ 4.5 TECU in 2005 to 8.9 ${\pm}$ 0.6 TECU in 2008, and the annual difference is constantly about 1.6 TECU, and increases to 10.3 ${\pm}$ 1.8 TECU in 2009. The annual variations of diurnal amplitude show the seasonal features that are scattered in summer and the enhancements near equinoxes are apparent in the whole years. In contrast, the semidiurnal amplitudes show the disturbed annual peaks in winter and its enhancements near equinoxes are unapparent. The diurnal phases are not constant in winter and show near 12 local time (LT). The semidiurnal phases have a seasonal pattern between 00 LT and 06 LT. Consequently, the KSJ GPS TEC variations show the significant semidiurnal variation in summer from December to February under the solar minimum between 2005 and 2009. The feature is considered as the Weddell Sea anomaly of larger nighttime electron density than a daytime electron density that has been observed around the Antarctica peninsula.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38C no.12
• /
• pp.1134-1149
• /
• 2013

Pub/sub system enables data users to access any necessary data without knowledge of the data producer and synchronization with the data producer. It is widely used as the middleware technology for the data-centric services. DDS (Data Distribution Service) is a standard middleware supported by the OMG (Object Management Group), one of global standardization organizations. It is considered quite useful as a standard middleware for US military services. However, it is well-known that it takes considerably long time in searching the Participants and Endpoints in the system, especially when the system is booting up. In this paper, we propose a discovery scheme to reduce the latency when the participants and Endpoints are densely distributed in a small area. We propose to modify the standard DDS discovery process in three folds. First, we integrate the Endpoint discovery process with the Participant discovery process. Second, we reduce the number of connections per participant during the discovery process by adopting the concept of successors in Distributed Hashing scheme. Third, instead of UDP, the participants are connected through TCP to exploit the reliable delivery feature of TCP. We evaluated the performance of our scheme by comparing with the standard DDS discovery process. The evaluation results show that our scheme achieves quite lower discovery latency in case that the Participants and the Endpoints are densely distributed in a local network.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.6
• /
• pp.39-47
• /
• 2010

A research of image-based articulated pose estimation has some problems such as detection of human feature, precise pose estimation, and real-time performance. In particular, various methods are currently presented for recovering many joints of human body. We propose the novel numerical inverse kinematics improved with the UKF(unscented Kalman filter) in order to estimate the human pose in real-time. An existing numerical inverse kinematics is required many iterations for solving the optimal estimation and has some problems such as the singularity of jacobian matrix and a local minima. To solve these problems, we combine the UKF as a tool for optimal state estimation with the numerical inverse kinematics. Combining the solution of the numerical inverse kinematics with the sampling based UKF provides the stability and rapid convergence to optimal estimate. In order to estimate the human pose, we extract the interesting human body using both background subtraction and skin color detection algorithm. We localize its 3D position with the camera geometry. Next, through we use the UKF based numerical inverse kinematics, we generate the intermediate joints that are not detect from the images. Proposed method complements the defect of numerical inverse kinematics such as a computational complexity and an accuracy of estimation.

• Journal of the Korean Society for Marine Environment & Energy
• /
• v.20 no.1
• /
• pp.55-61
• /
• 2017

This study conducted a survey to evaluate the disaster response capability of coastal residents and analyzed the characteristics. For the sampling process, nonrandom sampling method was used. Sample size is 4,520 and sample error is ${\pm}1.5%p$ at 95% confidence level. As a result of the survey, 72% and 68% of the respondents said that they recognized the emergency contact network and listened to the disaster broadcast. On the other hand, 17% and 18% said that they organized the local voluntary disaster prevention teams and participated in disaster preparedness training. In addition, male's disaster response capability was higher than female's, and first aid techniques and participation in disaster preparedness training were higher in teens and twenties. By occupation, public official possess the highest response capability. By region, it was high in the East coast and low in the South coast. It is necessary that the authorities improve the national disaster preparedness training and publicity to enhance the coastal disaster response capability of coastal residents.

• Journal of the Korean earth science society
• /
• v.36 no.3
• /
• pp.210-221
• /
• 2015

The terrain features and surface characteristics are the most important elements not only in meteorological modeling but also in air quality modeling. The diurnal evolution of local climate over complex terrain may be significantly controlled by the ground irregularities. Such topographic features can affect a thermally driven flow, either directly by causing changes in the wind direction or indirectly, by inducing significant variations in the ground temperature. Over a complex terrain, these variations are due to the nonuniform distribution of solar radiation, which is highly determined by the ground geometrical characteristics, i.e. slope and orientation. Therefore, the accuracy of prediction of regional scale circulation is strong associated with the accuracy of land-use and topographic information in meso-scale circulation assessment. The objective of this work is a numerical simulation using MM5-A2C model with the detailed topography and land-use information as the surface boundary conditions of the air flow field in mountain regions. Meteorological conditions estimated by MM5-A2C command a great influence on the dispersion of mountain areas with the reasonable feature of topography where there is an important difference in orographic forcing.

• Journal of Korean Society of Forest Science
• /
• v.105 no.1
• /
• pp.48-62
• /
• 2016

This study was conducted to classify forest vegetation structure and stand feature of Mt. Deogyusan National Park from Hyangjeukbong to Jungbong, 48 plots were surveyed. The type classification of the vegetation structure was performed with Z-M phytosociological method. As a result, Quercus mongolica community group was classified into the Picea jezoensis community, Carpinus cordata community and Tilia amurensis community in community unit. P. jezoensis community was subdivided into Deutzia glabrata group and Viburnum opulus var. calvescens group in group unit. D. glabrata group was subdivided into Acer mandshuricum subgroup and Ribes mandshuricum subgroup and V. opulus var. calvescens group was subdivided into Hemerocallis dumortieri subgroup and Prunus padus subgroup in subgroup unit. In the result of estimating the importance value, it constituted Q. mongolica (23.9%), Abies koreana (14.7%), Taxus cuspidata (10.2%), P. jezoensis (8.2%) and Betula ermanii (7.4%) in tree layer. It constituted Acer komarovii (18.6%), Acer pseudosieboldianum (18.4%) and Q. mongolica (8.9%) in subtree layer. It constituted Rhododendron schlippenbachii (20.7%), A. pseudosieboldianum (17.4%) and Symplocos chinensis (8.5%) in shrub layer. Indicator species analysis of vegetation unit 1 was consisted of Hydrangea serrata, Fraxinus mandshurica and D. glabrata that species prefer moist valley in subalpine or rocks. In the results of analyzing the species diversity, vegetation unit 1, 4 and 5 represented that there were different and complex local distributions. As in the similarity between the vegetation units, the vegetation units 1, 2, 3 and 4 represented high with 0.5 or above. It represented that there wasn't no differences on composition species in vegetation units.