• The Journal of the Korea Contents Association
• /
• v.20 no.7
• /
• pp.674-682
• /
• 2020

Recently, the problem of air pollution has become a global concern due to industrialization and overcrowding. Air pollution can cause various adverse effects on human health, among which respiratory diseases such as asthma, which have been of interest in this study, can be directly affected. Previous studies have used clinical data to identify how air pollutant affect diseases such as asthma based on relatively small samples. This is high likely to result in inconsistent results for each collection samples, and has significant limitations in that research is difficult for anyone other than the medical profession. In this study, the main focus was on predicting the actual asthmatic occurrence, based on data on the atmospheric environment data released by the government and the frequency of asthma outbreaks. First of all, this study verified the significant effects of each air pollutant with a time lag on the outbreak of asthma through the time-lag Pearson Correlation Coefficient. Second, train data built on the basis of verification results are utilized in Deep Learning algorithms, and models optimized for predicting the asthmatic occurrence are designed. The average error rate of the model was about 11.86%, indicating superior performance compared to other machine learning-based algorithms. The proposed model can be used for efficiency in the national insurance system and health budget management, and can also provide efficiency in the deployment and supply of medical personnel in hospitals. And it can also contribute to the promotion of national health through early warning of the risk of outbreak by atmospheric environment for chronic asthma patients.

• Convergence Security Journal
• /
• v.21 no.2
• /
• pp.57-66
• /
• 2021

Deep neural networks (DNNs) provide excellent performance for image, speech, and pattern recognition. However, DNNs sometimes misrecognize certain adversarial examples. An adversarial example is a sample that adds optimized noise to the original data, which makes the DNN erroneously misclassified, although there is nothing wrong with the human eye. Therefore studies on defense against adversarial example attacks are required. In this paper, we have experimentally analyzed the success rate of detection for adversarial examples by adjusting various parameters. The performance of the ensemble defense method was analyzed using fast gradient sign method, DeepFool method, Carlini & Wanger method, which are adversarial example attack methods. Moreover, we used MNIST as experimental data and Tensorflow as a machine learning library. As an experimental method, we carried out performance analysis based on three adversarial example attack methods, threshold, number of models, and random noise. As a result, when there were 7 models and a threshold of 1, the detection rate for adversarial example is 98.3%, and the accuracy of 99.2% of the original sample is maintained.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.43-45
• /
• 2021

The short-term quantitative precipitation prediction (QPF) system is important socially and economically to prevent damage from severe weather. Recently, many studies for short-term QPF model applying the Deep Neural Network (DNN) has been conducted. These studies require the sophisticated pre-processing because the mistreatment of various and vast meteorological data sets leads to lower performance of QPF. Especially, for more accurate prediction of the non-linear trends in precipitation, the dataset needs to be carefully handled based on the physical and dynamical understands the data. Thereby, this paper proposes the following approaches: i) refining and combining major factors (weather radar, terrain, air temperature, and so on) related to precipitation development in order to construct training data for pattern analysis of precipitation; ii) producing predicted precipitation fields based on Convolutional with ConvLSTM. The proposed algorithm was evaluated by rainfall events in 2020. It is outperformed in the magnitude and strength of precipitation, and clearly predicted non-linear pattern of precipitation. The algorithm can be useful as a forecasting tool for preventing severe weather.

• Journal of Broadcast Engineering
• /
• v.25 no.4
• /
• pp.569-575
• /
• 2020

The serial arc is one of factors causing electrical fires. Over past decades, various researches have been carried out to detect arc occurrences. Even though frequency analysis, wavelet, and statistical features have been used, additional steps such as transformation and feature extraction are required. On the contrary, deep learning models directly use the raw data without any feature extraction processes. Therefore, the usage of time-domain data is preferred, but the performance is not satisfactory. To solve this problem, subsequent 1-D signals are transformed into 2-D data that can feed into a convolutional neural network (CNN). Experiments validated that CNN model outperforms deep neural network (DNN) by the classification accuracy of 8.6%. In addition, data augmentation is utilized, resulting in the accuracy improvement by 14%.

• Journal of Broadcast Engineering
• /
• v.24 no.5
• /
• pp.735-742
• /
• 2019

In this paper, we propose a method of extracting speech features for phoneme recognition based on spikegram. The Fourier-transform-based features are widely used in phoneme recognition, but they are not extracted in a biologically plausible way and cannot have high temporal resolution due to the frame-based operation. For better phoneme recognition, therefore, it is desirable to have a new method of extracting speech features, which analyzes speech signal in high temporal resolution following the model of human auditory system. In this paper, we analyze speech signal based on a spikegram that models feature extraction and transmission in auditory system, and then propose a method of feature extraction from the spikegram for phoneme recognition. We evaluate the performance of proposed features by using a DNN-based phoneme recognizer and confirm that the proposed features provide better performance than the Fourier-transform-based features for short-length phonemes. From this result, we can verify the feasibility of new speech features extracted based on auditory model for phoneme recognition.

• Journal of Korea Society of Industrial Information Systems
• /
• v.23 no.6
• /
• pp.29-38
• /
• 2018

In this paper, we developed a virtual cloth try-on (VTON) technology that segement the cloth image worn on the mannequin and applies it to the user 's photograph. The two-dimensional image-based virtual wear study which does not require three-dimensional information of cloth and model is of practical value, but the research result shows that there are limitations of of the current technology for the problem of occlusion or distortion. In this study, we proposed an algorithm to apply the results obtained from the DNN- based segmentation and posture estimation to the user 's photograph, assuming that the mannequin cloth reduces the difficulties in this part. In order to improve the performance compared with the existing one, we used the validity check of the pre-attitude information, the improvement of the deformation using the outline, and the improvement of the divided area. As a result, a significantly improved result image of more than 50% was obtained.

• Phonetics and Speech Sciences
• /
• v.11 no.2
• /
• pp.57-64
• /
• 2019

In this paper, we construct a Korean text-to-speech system using the Merlin toolkit which is an open source system for speech synthesis. In the text-to-speech system, the HMM-based statistical parametric speech synthesis method is widely used, but it is known that the quality of synthesized speech is degraded due to limitations of the acoustic modeling scheme that includes context factors. In this paper, we propose an acoustic modeling architecture that uses deep neural network technique, which shows excellent performance in various fields. Fully connected deep feedforward neural network (DNN), recurrent neural network (RNN), gated recurrent unit (GRU), long short-term memory (LSTM), bidirectional LSTM (BLSTM) are included in the architecture. Experimental results have shown that the performance is improved by including sequence modeling in the architecture, and the architecture with LSTM or BLSTM shows the best performance. It has been also found that inclusion of delta and delta-delta components in the acoustic feature parameters is advantageous for performance improvement.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.12
• /
• pp.411-418
• /
• 2020

In the field of information security, IDS(Intrusion Detection System) is normally classified in two different categories: signature-based IDS and anomaly-based IDS. Many studies in anomaly-based IDS have been conducted that analyze network traffic data generated in cyberspace by machine learning algorithms. In this paper, we studied pre-processing methods to overcome performance degradation problems cashed by rare classes. We experimented classification performance of a Machine Learning algorithm by reconstructing data set based on rare classes and semi rare classes. After reconstructing data into three different sets, wrapper and filter feature selection methods are applied continuously. Each data set is regularized by a quantile scaler. Depp neural network model is used for learning and validation. The evaluation results are compared by true positive values and false negative values. We acquired improved classification performances on all of three data sets.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.18 no.4
• /
• pp.44-57
• /
• 2019

Various studies have been conducted to solve traffic congestions in many metropolitan cities through accurate traffic flow prediction. Most studies are based on the assumption that past traffic patterns repeat in the future. Models based on such an assumption fall short in case irregular traffic patterns abruptly occur. Instead, the approaches such as predicting traffic pattern through big data analytics and artificial intelligence have emerged. Specifically, deep learning algorithms such as RNN have been prevalent for tackling the problems of predicting temporal traffic flow as a time series. However, these algorithms do not perform well in terms of long-term prediction. In this paper, we take into account various external factors that may affect the traffic flows. We model the correlation between the multi-dimensional context information with temporal traffic speed pattern using deep neural networks. Our model trained with the traffic data from TOPIS system by Seoul, Korea can predict traffic speed on a specific date with the accuracy reaching nearly 90%. We expect that the accuracy can be improved further by taking into account additional factors such as accidents and constructions for the prediction.