• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.126-128
• /
• 2022

경량화 기법 중 하나인 Knowledge distillation 은 최근 object detection task 에 적용되고 있다. Knowledge distillation 은 3 가지 범주로 나뉘는데 그들 중에서 Self-Knowledge distillation 은 기존의 Knowledge distillation 에서의 pre-trained teacher 에 대한 의존성 문제를 완화시켜준다. Self-Knowledge distillation 또한 object detection task 에 적용되어 training cost 를 줄이고 고전적인 teacher-based methods 보다 좋은 성능을 성취했다.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.4
• /
• pp.691-708
• /
• 2022

As the technology of machine learning and deep learning became common, it began to be applied to research on anomaly(abnormal) detection of industrial control systems. In Korea, the HAI dataset was developed and published to activate artificial intelligence research for abnormal detection of industrial control systems, and an AI contest for detecting industrial control system security threats is being conducted. Most of the anomaly detection studies have been to create a learning model with improved performance through the ensemble model method, which is applied either by modifying the existing deep learning algorithm or by applying it together with other algorithms. In this study, a study was conducted to improve the performance of anomaly detection with a post-processing method that detects abnormal data and corrects the labeling results, rather than the learning algorithm and data pre-processing process. Results It was confirmed that the results were improved by about 10% or more compared to the anomaly detection performance of the existing model.

• Journal of Oral Medicine and Pain
• /
• v.32 no.4
• /
• pp.347-355
• /
• 2007

The aim of this study was to evaluate the differences of salivary lead (Pb) and cadmium (Cd) concentrations, using ASV analysis, after various pre-treatment procedures. 10 unstimulated whole saliva samples of non-exposed subjects to Pb and Cd were collected. Each sample was divided into 6 aliquots and centrifugation was performed in only 3 aliquots. After centrifugation, 3 different types of pre-treatment procedures were carried out. Also, these pre-treatment procedures were carried out for another 3 aliquots, without centrifugation. Pre-treated aliquots were analyzed electrochemically, by ASV. The results are as follows: 1. Mean concentration of Pb in saliva after centrifugation was significantly higher than that of non-centrifugation. 2. In the detection sensitivity of Pb in saliva, those of simple dilution technique by HCl and acid digestion technique by nitric acid were significantly higher than that of simple dilution technique by electrolyte. 3. Mean concentration of Cd in saliva after centrifugation was significantly higher than that of non-centrifugation. 4. In the detection sensitivity of Cd in saliva, those of simple dilution technique by HCl and acid digestion technique by nitric acid were higher than that of simple dilution technique by electrolyte. But, there were no significant differences between them.

• Journal of Korea Multimedia Society
• /
• v.19 no.3
• /
• pp.633-641
• /
• 2016

Automatic defect inspection system is composed of the step in the pre-processing, defect candidate detection, and classification. Polarizing films containing various defects should be minimized over-detection for classifying defect blobs. In this paper, we propose a defect detection algorithm using a skewness of histogram for minimizing over-detection. In order to detect up defects with similar to background pixel, we are used the characteristics of the local region. And the real defect pixels are distinguished from the noise using the probability density function. Experimental results demonstrated the minimized over-detection by utilizing the artificial images and real polarizing film images.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.4
• /
• pp.249-255
• /
• 2022

Potholes are an important clue to the structural defects of asphalt pavement and cause many casualties and property damage. Therefore, accurate pothole detection is an important task in road surface maintenance. Many machine learning technologies are being introduced for pothole detection, and data preprocessing is required to increase the efficiency of deep learning models. In this paper, we propose a preprocessing method that emphasizes important textures and shapes in pothole datasets. The proposed preprocessing method uses intensity transformation to reduce unnecessary elements of the road and emphasize the texture and shape of the pothole. In addition, the feature of the porthole is detected using Superpixel and Sobel edge detection. Through performance comparison between the proposed preprocessing method and the existing preprocessing method, it is shown that the proposed preprocessing method is a more effective method than the existing method in detecting potholes.

• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.9
• /
• pp.699-704
• /
• 2016

In this paper, we propose a development method for a drowsiness detection system using retinex theory and edge information for vehicle safety. Detection of a drowsy state of a driver is very important because the drowsiness of driver is often the main cause of many car accidents. After acquiring an image of the entire face, we executed the pre-process step using the retinex theory. We then applied a technique for the detection of the white pixels using edge information. Experimental results showed that the proposed method improved the accuracy of detecting drowsiness to nearly 98%, and can be used to prevent a car accident caused by the driver's drowsiness.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1995.10a
• /
• pp.87-94
• /
• 1995

In highway bridges robust damage detection exercises are mandatory to secure the safety of the structures from hostile environmental conditions such as fatigue earthquake, wind, and corrosion. This paper presents a damage detection practice in a full-scale highway bridge by utilizing modal response parameters of as-built and damaged states of the structure. first the test structure is described and modal testing procedures are outlined. Next, a damage detection model which yields information on the location of damage directly from changes in mode shapes is outlined. Finally, the damage detection model is implemented to predict the location of damage in the ten structure. From the results, it was found that the damage detection model accurately locates damage in the test structures for which modal parameters of only a single mode are available for pre-damage (as-built) and post-damage stages.

• Journal of Broadcast Engineering
• /
• v.6 no.2
• /
• pp.139-147
• /
• 2001

Since various noises degrade not only image quality but also compression efficiency in MPEG and H.263, pre-processing is necessary to reduce spatial and temporal noise and to increase ceding efficiency as well. In this paper, we propose a simplified method for noise detection, spatial and temporal noise reduction. Noise detection is based on correlation of the current pixel with its neighboring 4 pixels. Spatial noose reduction utilizes a non-rectangular median filter that is less complex than the conventional rectangular median filter. The proposed temporal filter is an IIR average filter using LUT(Look-up Table) to enhance subjective video quality. The proposed pre-processing method is very simple and efficient.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.623-625
• /
• 2021

Rubber o-rings are produced by conventional injection molding methods. In this case, products that are not normally molded are determined to be defective. However, if images acquired during image-based reading are read as original, there is a problem of poor accuracy. We have thus learned from convolutional neural networks using adaptive binarization and Convex Hull algorithms by extracting only rubber oring parts from the original images through pre-processing. During the test process, it was confirmed that the defect detection performance of the learning method applied pre-processing was better than the standard suggested.

• Phonetics and Speech Sciences
• /
• v.1 no.2
• /
• pp.85-94
• /
• 2009

The purpose of this study was to see the effects of pitch and length patterns on spoken word recognition. In Experiment 1, a syllable monitoring task was used to see the effects of pitch and length on the pre-lexical level of spoken word recognition. For both Seoul dialect speakers and Daegu dialect speakers, pitch and length did not affect the syllable detection processes. This result implies that there is little effect of pitch and length in pre-lexical processing. In Experiment 2, a lexical decision task was used to see the effect of pitch and length on the lexical access level of spoken word recognition. In this experiment, word frequency (low and high) as well as pitch and length was manipulated. The results showed that pitch and length information did not play an important role for Seoul dialect speakers, but that it did affect lexical decision processing for Daegu dialect speakers. Pitch and length seem to affect lexical access during the word recognition process of Daegu dialect speakers.