• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.368-369
• /
• 2011

This article suggests an online-based remote fault current mode discrimination method in order to identify the causes of the power line faults with various causes. For that, it refers to existing cause identification methods and categorizes modes by fault causes based on statistical techniques beforehand and performs the pretreatment process of fault currents by each cause acquired from the fault recorder into a topological plane in order to extract the characteristics of fault currents by each cause. After that, for the fault mode categorization, it discriminates modes by each cause using data by each cause as leaning data through utilizing RBF network based on FCM allowing self-organization in deciding the middle layer. And then it tests the validity of the suggested method as applying it to the data of the actual fault currents acquired from the fault recorder in the electric power transmission center.

• Journal of Digital Contents Society
• /
• v.2 no.1
• /
• pp.73-79
• /
• 2001

Most of the existing algorithms try to disseminate the multimedia contents of internet service provider(ISP), without taking into account the needs and characteristics of specific websites including e-learning systems with web-based .educational contents. Sometimes the client must select the best one among the replicated repositories. However, this is a less reliable approach because clients' selections are made without prior information on server load capacity. In this paper we propose an agent system inspired by the need of improving QoS of delivering web-based educational multimedia contents without incurring long access delays. This agent system consists of three components, Analyzer, Knowledge Base, and Automaton embedded the capacity algorithm. It analyzes and investigates traffic information collected from individual replicated server by learners' requests, and selects a server which is available and is expected to provide the fastest latency time and the lowest loaded capacity, and achieves high performance by dynamic replicating web resources among multiple repositories.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.22 no.5
• /
• pp.515-528
• /
• 2020

Underground concrete structures are usually designed to be used for decades, but in recent years, many of them are nearing their original life expectancy. As a result, it is necessary to promptly inspect and repair the structure, since it can cause lost of fundamental functions and bring unexpected problems. Therefore, personnel-based inspections and repairs have been underway for maintenance of underground structures, but nowadays, objective inspection technologies have been actively developed through the fusion of deep learning and image process. In particular, various researches have been conducted on developing a concrete crack detection algorithm based on supervised learning. Most of these studies requires a large amount of image data, especially, label images. In order to secure those images, it takes a lot of time and labor in reality. To resolve this problem, we introduce a method to increase the accuracy of crack area detection, improved by 0.25% on average by applying adversarial learning in this paper. The adversarial learning consists of a segmentation neural network and a discriminator neural network, and it is an algorithm that improves recognition performance by generating a virtual label image in a competitive structure. In this study, an efficient deep neural network learning method was proposed using this method, and it is expected to be used for accurate crack detection in the future.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.5C
• /
• pp.486-492
• /
• 2003

The vigilance threshold of conventional fuzzy ART algorithm decide whether to permit the mismatch between any input pattern and stored pattern. If the vigilance threshold was large, despite of little difference among input and stored patterns, the input pattern may be classified to new category. On the other hand, if the vigilance threshold was small, the similarity between two patterns may be accepted in spite of lots of difference and the input pattern are classified to category of the stored pattern. Therefore, the vigilance threshold for the image recognition must be experientially set for the good result. Moreover, it may occur in the fuzzy ART algorithm that the information of stored patterns is lost in the weight-adjusting process and the rate of pattern recognition is dropped. In this paper, I proposed the enhanced fuzzy ART algorithm that supports the dynamical setting of the vigilance threshold using the generalized intersection operator of fuzzy logic and the weight value being adaptively set in proportional to the current weight change and the previous weight by reflecting the frequency of the selection of winner node. For the performance evaluation of the proposed method, we applied to the recognition of container identifiers from shipping container images. The experiment showed that the proposed method produced fewer clusters than conventional ART2 and fuzzy ART algorithm. and had tile higher recognition rate.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.5
• /
• pp.437-443
• /
• 2015

This paper presents face recognition based on the fusion of visible image and thermal infrared (IR) texture estimated from the face image in the visible spectrum. The proposed face recognition scheme uses a multi- layer neural network to estimate thermal texture from visible imagery. In the training process, a set of visible and thermal IR image pairs are used to determine the parameters of the neural network to learn a complex mapping from a visible image to its thermal texture in the low-dimensional feature space. The trained neural network estimates the principal components of the thermal texture corresponding to the input visible image. Extensive experiments on face recognition were performed using two popular face recognition algorithms, Eigenfaces and Fisherfaces for NIST/Equinox database for benchmarking. The fusion of visible image and thermal IR texture demonstrated improved face recognition accuracies over conventional face recognition in terms of receiver operating characteristics (ROC) as well as first matching performances.

• Journal of KIISE:Computing Practices and Letters
• /
• v.7 no.6
• /
• pp.684-695
• /
• 2001

Recent recommender system uses a method of combining collaborative filtering system and content based filtering system in order to solve sparsity and first rater problem in collaborative filtering system. Collaborative filtering systems use a database about user preferences to predict additional topics. Content based filtering systems provide recommendations by matching user interests with topic attributes. In this paper, we describe a method for discovery of user preference through combining two techniques for recommendation that allows the application of machine learning algorithm. The proposed collaborative filtering method clusters user using genetic algorithm based on items categorized by Naive Bayes classifier and the content based filtering method builds user profile through extracting user interest using relevance feedback. We evaluate our method on a large database of user ratings for web document and it significantly outperforms previously proposed methods.

• KIPS Transactions on Software and Data Engineering
• /
• v.3 no.12
• /
• pp.511-522
• /
• 2014

During the development of the software, a variety of bugs are reported. Several bug tracking systems, such as, Bugzilla, MantisBT, Trac, JIRA, are used to deal with reported bug information in many open source development projects. Bug reports in bug tracking system would be triaged to manage bugs and determine developer who is responsible for resolving the bug report. As the size of the software is increasingly growing and bug reports tend to be duplicated, bug triage becomes more and more complex and difficult. In this paper, we present an approach to assign bug reports to appropriate developers, which is a main part of bug triage task. At first, words which have been included the resolved bug reports are classified according to each developer. Second, words in newly bug reports are selected. After first and second steps, vectors whose items are the selected words are generated. At the third step, TF-IDF(Term frequency - Inverse document frequency) of the each selected words are computed, which is the weight value of each vector item. Finally, the developers are recommended based on the similarity between the developer's word vector and the vector of new bug report. We conducted an experiment on Eclipse JDT and CDT project to show the applicability of the proposed approach. We also compared the proposed approach with an existing study which is based on machine learning. The experimental results show that the proposed approach is superior to existing method.

• Journal of the Korea Society for Simulation
• /
• v.20 no.4
• /
• pp.115-125
• /
• 2011

Manufacturing companies usually manage the process to achieve high quality using various types of control chart in statistical process control. When an assignable cause occurs in a process, the data in the control chart changes with different patterns by the specific causes. It is important in process control to classify the CCP (Control Chart Pattern) recognition for fast decision making. In former research, gathered data from process used to apply as raw data, leads to degrade the performance of recognizer and to decrease the learning speed. Therefore, feature based recognizer, employing feature extraction method, has been studied to enhance the classification accuracy and to reduce the dimension of data. We propose the method to extract features that take the distances between CCP data and reference vector generated from BDK (Bi-Directional Kohonen Network). We utilize those features as the input vectors in ANN (Artificial Neural Network) and compare with raw data applied ANN to evaluate the performance.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.1
• /
• pp.17-22
• /
• 2024

This paper introduces the implementation of a secondhand clothing trading system equipped with virtual fitting functionality based on deep learning. The proposed system provides users with the ability to visually try on secondhand clothing items online and assess their fit. To achieve this, it utilizes the Convolutional Neural Network (CNN) algorithm to create virtual representations of users considering their body shape and the design of the clothing. This enables buyers to pre-assess the fit of clothing items online before actually wearing them, thereby aiding in their purchase decisions. Additionally, sellers can present accurate clothing sizes and fits through the system, enhancing customer satisfaction. This paper delves into the CNN model's training process, system implementation, user feedback, and validates the effectiveness of the proposed system through experimental results.

• Journal of Convergence for Information Technology
• /
• v.11 no.6
• /
• pp.33-39
• /
• 2021

Speech recognition technology is being combined with deep learning and is developing at a rapid pace. In particular, voice recognition services are connected to various devices such as artificial intelligence speakers, vehicle voice recognition, and smartphones, and voice recognition technology is being used in various places, not in specific areas of the industry. In this situation, research to meet high expectations for the technology is also being actively conducted. Among them, in the field of natural language processing (NLP), there is a need for research in the field of removing ambient noise or unnecessary voice signals that have a great influence on the speech recognition recognition rate. Many domestic and foreign companies are already using the latest AI technology for such research. Among them, research using a convolutional neural network algorithm (CNN) is being actively conducted. The purpose of this study is to determine the non-voice section from the user's speech section through the convolutional neural network. It collects the voice files (wav) of 5 speakers to generate learning data, and utilizes the convolutional neural network to determine the speech section and the non-voice section. A classification model for discriminating speech sections was created. Afterwards, an experiment was conducted to detect the non-speech section through the generated model, and as a result, an accuracy of 94% was obtained.