• ETRI Journal
• /
• v.39 no.6
• /
• pp.841-850
• /
• 2017

We investigate neural network image reconstruction for magnetic particle imaging. The network performance strongly depends on the convolution effects of the spectrum input data. The larger convolution effect appearing at a relatively smaller nanoparticle size obstructs the network training. The trained single-layer network reveals the weighting matrix consisting of a basis vector in the form of Chebyshev polynomials of the second kind. The weighting matrix corresponds to an inverse system matrix, where an incoherency of basis vectors due to low convolution effects, as well as a nonlinear activation function, plays a key role in retrieving the matrix elements. Test images are well reconstructed through trained networks having an inverse kernel matrix. We also confirm that a multi-layer network with one hidden layer improves the performance. Based on the results, a neural network architecture overcoming the low incoherence of the inverse kernel through the classification property is expected to become a better tool for image reconstruction.

• Computers and Concrete
• /
• v.24 no.6
• /
• pp.555-560
• /
• 2019

During the application of shotcrete, a part of the concrete bounces back after hitting to the surface, the reinforcement or previously sprayed concrete. This rebound material is definitely not added to the mixture and considered as waste. In this study, a deep neural network model was developed to predict the rebound material during shotcrete application. The factors affecting rebound and the datasets of these parameters were obtained from previous experiments. The Long Short-Term Memory (LSTM) architecture of the proposed deep neural network model was used in accordance with this data set. In the development of the proposed four-tier prediction model, the dataset was divided into 90% training and 10% test. The deep neural network was modeled with 11 dependents 1 independent data by determining the most appropriate hyper parameter values for prediction. Accuracy and error performance in success performance of LSTM model were evaluated over MSE and RMSE. A success of 93.2% was achieved at the end of training of the model and a success of 85.6% in the test. There was a difference of 7.6% between training and test. In the following stage, it is aimed to increase the success rate of the model by increasing the number of data in the data set with synthetic and experimental data. In addition, it is thought that prediction of the amount of rebound during dry-mix shotcrete application will provide economic gain as well as contributing to environmental protection.

• Journal of Information Technology Services
• /
• v.18 no.4
• /
• pp.83-95
• /
• 2019

It is becoming more and more difficult to cope with new knowledge and technology required by society by the efforts of one person or organization according to the development of science and technology. As a method to overcome this, collaborative research is becoming important. This tendency is increasing in the government R&D projects as well, and the 'A' test research institute, which is the subject of this paper, is also increasing a collaborative research. The purpose of this study is to analyze the network characteristics among the participating researchers in the government R&D project conducted by the institution A, and to ascertain how the network characters of the researchers actually affect the financial performance of the team. The results of the analysis show that 'closeness centrality' and 'degree of centrality' contribute positively to the financial performance of the team. On the other hand, 'betweenness centrality' and 'eigenvector centrality' have a negative effect on the financial performance of the team because they are not directly related to financial performance.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.1E
• /
• pp.54-65
• /
• 1998

This study attempts to test the classifying performance of a neural network and thereby examine its applicability to the signals distorted in a shallow water environment. Linear frequency modulated(LFM) signals are simulated by using an acoustic model and also measured through sea experiment. The network is constructed to have three layers and trained on both data sets. To get normalized power spectra as feature vectors, the study considers the three transforms : shot-time Fourier transform (STFT), wavelet transform (WT) and pseudo Wigner-Ville distribution (PWVD). After trained on the simulated signals over water depth, the network gives over 95% performance with the signal to noise ratio (SNR) being up to-10 dB. Among the transforms, the PWVD presents the best performance particularly in a highly noisy condition. The network performs worse with the summer sound speed profile than with the winter profile. It is also expected to present much different performance by the variation of bottom property. When the network is trained on the measured signals, it gives a little better results than that trained on the simulated data. In conclusion, the simulated signals are successfully applied to training a network, and the trained network performs well in classifying the signals distorted by a surrounding environment and corrupted by noise.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.5
• /
• pp.1198-1206
• /
• 2011

In this paper, we proposed a home network integrated dimmer switch using Z-wave Mesh Network. The dimmer switch come with wired and wireless network-based intelligent lighting switch that can be installed in a apartments with home network technologies. This switch is ease to install and operate. Furthermore, it provide power saving through efficient control of light using motion detect sensor and light intensity sensor. In order to evaluate the performance of our developed home network integrated dimmer switch, we performed a load performance test and from our result, it showed excellent performance in terms of lighting control and communication speed.

• Imaging Science in Dentistry
• /
• v.52 no.2
• /
• pp.219-224
• /
• 2022

Purpose: This study aimed to evaluate the performance of transfer learning in a deep convolutional neural network for classifying implant fixtures. Materials and Methods: Periapical radiographs of implant fixtures obtained using the Superline (Dentium Co. Ltd., Seoul, Korea), TS III(Osstem Implant Co. Ltd., Seoul, Korea), and Bone Level Implant(Institut Straumann AG, Basel, Switzerland) systems were selected from patients who underwent dental implant treatment. All 355 implant fixtures comprised the total dataset and were annotated with the name of the system. The total dataset was split into a training dataset and a test dataset at a ratio of 8 to 2, respectively. YOLOv3 (You Only Look Once version 3, available at https://pjreddie.com/darknet/yolo/), a deep convolutional neural network that has been pretrained with a large image dataset of objects, was used to train the model to classify fixtures in periapical images, in a process called transfer learning. This network was trained with the training dataset for 100, 200, and 300 epochs. Using the test dataset, the performance of the network was evaluated in terms of sensitivity, specificity, and accuracy. Results: When YOLOv3 was trained for 200 epochs, the sensitivity, specificity, accuracy, and confidence score were the highest for all systems, with overall results of 94.4%, 97.9%, 96.7%, and 0.75, respectively. The network showed the best performance in classifying Bone Level Implant fixtures, with 100.0% sensitivity, specificity, and accuracy. Conclusion: Through transfer learning, high performance could be achieved with YOLOv3, even using a small amount of data.

• Proceedings of the KSR Conference
• /
• 2008.06a
• /
• pp.1887-1891
• /
• 2008

In this paper, we dealt with a partial discharge (PD) measurement method that has been accepted as an effective and non-destructive technique to estimate insulation performance of low-voltage induction motors. The PD measurement system consists of a coupling network, a low noise amplifier, and associated electronics. A shielded box was used to reduce environmental noise. Frequency characteristic of the coupling network was estimated by a sinusoidal signal input, and the low cut-off frequency of the coupling network was 1 MHz (-3 dB). Also, we carried out a calibration test for the PD measurement system. Sensitivity of the system was of 84 m$V_{max}$/pC between stator winding and enclosure. In application test on a low-voltage three phase induction motor (5 HP), we could detect 88 pC at AC 800 $V_{max}$.

• Proceedings of the KIEE Conference
• /
• 2005.07d
• /
• pp.3040-3042
• /
• 2005

During the last several years, interest in broad-band power line communications (PLC) has been grown over medium voltage(MV) power distribution lines as well as low voltage lines. This paper introduces a medium voltage PLC test field that is set up in the suburbs of Euiwang city in Korea. This test field could be used not only for the measurement of communication channel environment but also for internet service. This paper shows the configuration of medium voltage test field with network devices like MV signal coupler and the results of channel environment like noise and impulse response. It also shows the service performance of PLC access network through network management system.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.15 no.3
• /
• pp.39-44
• /
• 2006

In this study, we nodestructive test based on ultrasonic test as inspection method and compared backpropagation neural network(BPNN) with probabilistic neural network(PNN) as pattern recognition algorithm of welding flasw. For this purpose, variables are applied the same to two algorithms. Where, feature variables are zooming flaw signals of reflected whole signals from welding flaws in time domain. Through this process, we confirmed advantages/disadvantages of two algorithms and identified application methods of two algorithms.

• Journal of Korea Society of Industrial Information Systems
• /
• v.22 no.5
• /
• pp.17-24
• /
• 2017

Recently Deep-learning, Technologies using Large or Deep Artificial Neural Networks, have Shown Remarkable Performance, and the Increasing Size of the Network Contributes to its Performance Improvement. However, the Increase in the Size of the Neural Network Leads to an Increase in the Calculation Amount, which Causes Problems Such as Circuit Complexity, Price, Heat Generation, and Real-time Restriction. In This Paper, We Propose and Test a Method to Reduce the Number of Network Connections by Effectively Pruning the Redundancy in the Connection and Showing the Difference between the Performance and the Desired Range of the Original Neural Network. In Particular, we Proposed a Simple Method to Improve the Performance by Re-learning and to Guarantee the Desired Performance by Allocating the Error Rate per Layer in Order to Consider the Difference of each Layer. Experiments have been Performed on a Typical Neural Network Structure such as FCN (full connection network) and CNN (convolution neural network) Structure and Confirmed that the Performance Similar to that of the Original Neural Network can be Obtained by Only about 1/10 Connection.