• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.29 no.1
• /
• pp.117-125
• /
• 2019

Deep Learning, which is being used in various fields recently, is being threatened with Adversarial Attack. In this paper, we experimentally verify that the classification accuracy is lowered by adversarial samples generated by malicious attackers in image classification models. We used MNIST dataset and measured the detection accuracy by injecting adversarial samples into the Autoencoder classification model and the CNN (Convolution neural network) classification model, which are created using the Tensorflow library and the Pytorch library. Adversarial samples were generated by transforming MNIST test dataset with JSMA(Jacobian-based Saliency Map Attack) and FGSM(Fast Gradient Sign Method). When injected into the classification model, detection accuracy decreased by at least 21.82% up to 39.08%.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.29-32
• /
• 2020

Due to the lack of improper image acquisition process, noise induction is an inevitable step. As a result, objective image quality assessment (IQA) plays an important role in estimating the visual quality of noisy image. Plenty of IQA methods have been proposed including traditional signal processing based methods as well as current deep learning based methods where the later one shows promising performance due to their complex representation ability. The deep learning based methods consists of several convolution layers and down sampling layers for feature extraction and fully connected layers for regression. Usually, the down sampling is performed by using max-pooling layer after each convolutional block. We reveal that this max-pooling causes information loss despite of knowing their importance. Consequently, we propose a better IQA method that replaces the max-pooling layers with strided convolutions to down sample the feature space and since the strided convolution layers have learnable parameters, they preserve optimal features and discard redundant information, thereby improve the prediction accuracy. The experimental results verify the effectiveness of the proposed method.

• 한국해양학회지
• /
• v.28 no.3
• /
• pp.192-201
• /
• 1993

To reduce tsunami hazards, it is necessary to develope the methods which can simulate tsunami wave signals of coastal areas. In the present paper, it is attempted t use Sign Method for analyzing and simulating recorded tsunami signals. A tsunami record Y(t) can be represented as the convolution integral of a source evolution function E(t') and a wave propagation function K(t-t') Y(t)=.int. E(t')K(t-t')dt' A source function contains the peculiarities of a tsunami generator. A wave function is a kind of transfer function which contains the characteristics of a wave propagation path. The source functions and the wave function and the wave functions of 9 Korean coast points and 6 Japan coast points are estimated, and the tsunami wave signals are simulated by the convolution integrals of the source functions and the wave functions. According to the results of analysis, the Sign Method is an effective method for simulating tsunami wave signals of Korean coast points which are located far from tsunami source areas. Furthermore, if the source function of a neighboring point ad the wave function of an another tsunami are given, unrecorded tsunami wave also can be estimated.

• Structural Engineering and Mechanics
• /
• v.29 no.2
• /
• pp.237-240
• /
• 2008

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.6
• /
• pp.163-171
• /
• 2008

In this paper, we present an adaptive parametric cubic convolution technique in order to enlarge the low resolution image to the high resolution image. The proposed method consists of two steps. During the first interpolation step, we acquire adaptive parameters in introducing a new cost-function to reflect frequency properties. And, the second interpolation step performs cubic convolution by applying the parameters obtained from the first step. The enhanced interpolation kernel using adaptive parameters produces output image better than the conventional one using a fixed parameter. Experimental results show that the proposed method can not only provides the performances of $0.5{\sim}4dB$ improvements in terms of PSNR, but also exhibit better edge preservation ability and original image similarity than conventional methods in the enlarged images.

• Journal of IKEEE
• /
• v.23 no.1
• /
• pp.134-142
• /
• 2019

As the development of internet of energy (IoE) technologies and spread of various electronic devices have diversified patterns of energy consumption, the reliability of demand prediction has decreased, causing problems in optimization of power generation and stabilization of power supply. In this study, we propose a deep learning method, 1-Dimention-Convolution and Bidirectional Long Short-Term Memory (1D-ConvBLSTM), that combines a convolution neural network (CNN) and a Bidirectional Long Short-Term Memory(BLSTM) for highly reliable demand forecasting by effectively extracting the energy consumption pattern. In experimental results, the demand is predicted with the proposed deep learning method for various number of learning iterations and feature maps, and it is verified that the test data is predicted with a small number of iterations.

• Journal of Broadcast Engineering
• /
• v.27 no.6
• /
• pp.885-896
• /
• 2022

Motor failure in manufacturing plays an important role in future A/S and reliability. Motor failure is detected by measuring sound, current, and vibration. For the data used in this paper, the sound of the car's side mirror motor gear box was used. Motor sound consists of three classes. Sound data is input to the network model through a conversion process through MelSpectrogram. In this paper, various methods were applied, such as data augmentation to improve the performance of classifying fault motors and various methods according to class imbalance were applied resampling, reweighting adjustment, change of loss function and representation learning and classification into two stages. In addition, the curriculum learning method and self-space learning method were compared through a total of five network models such as Bidirectional LSTM Attention, Convolutional Recurrent Neural Network, Multi-Head Attention, Bidirectional Temporal Convolution Network, and Convolution Neural Network, and the optimal configuration was found for motor sound classification.

• Proceedings of the Korea Concrete Institute Conference
• /
• 1999.10a
• /
• pp.713-716
• /
• 1999

The presence of voids under pavements or behind tunnel linings results in the deterioration. One method of detecting such voids by non-destructive means is radar. This research is devoted to quantitatively evaluating the efficiency of such non-destructive tests with radar. As a foundation to this ongoing research, which aims to estimate the thickness of voids using radar, an analysis method based on radar signal processing using convolution technique is carried out with various void thicknesses in embedded layer which has different electromagnetic properties. The computed results were verified by comparing the test results. As a result, a proposed method in this study has a possibilty of estimating the thickness of voids with good accuracy.

• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.2060-2063
• /
• 2003

Terrestrial DMB(Digital Multimedia Broad-casting) system that is now under standardization in Korea offers multimedia broadcasting services at mobile environment and is based on Eureka-147 DAB(Digital Audio Broadcasting) for transmission method. Also DMB provides the error protection method of convolution coding. In this paper, we study on the effective error resilience coding of MPEG-4 video stream over DMB system. In our algorithm, the first, we partition the MPEG-4 data using the MPEG-4 data partitioning method, and then controls the convolution coding rate according to the importance of the partitioned data. From our simulation result, we show that our algorithm is proper for terrestrial DMB services.

• Journal of IKEEE
• /
• v.24 no.3
• /
• pp.883-889
• /
• 2020

In this paper, we propose a method to use PV(Park's Vector) pattern for inductive motor stator fault diagnosis using CNN(Convolution Neural Network). The conventional CNN based fault diagnosis method was performed by imaging three-phase currents, but this method was troublesome to perform normalization by artificially setting the starting point and phase of current. However, when using PV pattern, the problem of normalization could be solved because the 3-phase current shows a certain circular pattern. In addition, the proposed method is proved to be superior in the accuracy of CNN by 18.18[%] compared to the previous current data image due to the autonomic normalization.