• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.11
• /
• pp.455-464
• /
• 2022

There are two unique characteristics of the datasets from a manufacturing process. They are the severe class imbalance and lots of Out-of-Distribution samples. Some good strategies such as the oversampling over the minority class, and the down-sampling over the majority class, are well known to handle the class imbalance. In addition, SMOTE has been chosen to address the issue recently. But, Out-of-Distribution samples have been studied just with neural networks. It seems to be hardly shown that Out-of-Distribution detection is applied to the predictive model using conventional machine learning algorithms such as SVM, Random Forest and KNN. It is known that conventional machine learning algorithms are much better than neural networks in prediction performance, because neural networks are vulnerable to over-fitting and requires much bigger dataset than conventional machine learning algorithms does. So, we suggests a new approach to utilize Out-of-Distribution detection based on SVM algorithm. In addition to that, bagging technique will be adopted to improve the precision of the model.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.12
• /
• pp.1535-1541
• /
• 2019

To carry out safe and rapid decontamination in radiological accident areas, acquisition of various information on radiation sources is needed. In particular, to figure out the location and distribution of radiation sources is essential for rapid follow-up and removal of contaminants as well as minimizing worker damage. The radiation distribution detection device is used to obtain the position and distribution information of the radiation source. In the case of a radiation distribution detection device, a detection sensor unit is generally composed of a single sensor, and the detection range is limited due to the physical characteristics of the single sensor. We applied a calibration detector for controlling the detection sensitivity of a single sensor for radiation detection and improved the limited detection range of radiation dose rate. Also, gamma irradiation test confirmed the improvement of radiation distribution detection range.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.9
• /
• pp.2927-2941
• /
• 2022

The primary task of machine fault diagnosis is to judge whether the current state is normal or damaged, so it is a typical binary classification problem with mutual exclusion. Mutually exclusive events and out-of-domain detection have one thing in common: there are two types of data and no intersection. We proposed a fusion model method to improve the accuracy of machine fault diagnosis, which is based on the mutual exclusivity of events and the commonality of out-of-distribution detection, and finally generalized to all binary classification problems. It is reported that the performance of a convolutional neural network (CNN) will decrease as the recognition type increases, so the variational auto-encoder (VAE) is used as the primary model. Two VAE models are used to train the machine's normal and fault sound data. Two reconstruction probabilities will be obtained during the test. The smaller value is transformed into a correction value of another value according to the mutually exclusive characteristics. Finally, the classification result is obtained according to the fusion algorithm. Filtering normal data features from fault data features is proposed, which shields the interference and makes the fault features more prominent. We confirm that good performance improvements have been achieved in the machine fault detection data set, and the results are better than most mainstream models.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.16 no.1
• /
• pp.1-8
• /
• 2021

Detecting Out-of-Distribution (OOD) data is fundamentally required when Deep Neural Network (DNN) is applied to real-world AI such as autonomous driving. However, modern DNNs are quite vulnerable to the over-confidence problem even if the test data are far away from the trained data distribution. To solve the problem, this paper proposes a novel Adversarial-Mixup training method to let the DNN model be more robust by detecting OOD data effectively. Experimental results show that the proposed Adversarial-Mixup method improves the overall performance of OOD detection by 78% comparing with the State-of-the-Art methods. Furthermore, we show that the proposed method can alleviate the over-confidence problem by reducing the confidence score of OOD data than the previous methods, resulting in more reliable and robust DNNs.

• Asia pacific journal of information systems
• /
• v.31 no.3
• /
• pp.335-357
• /
• 2021

As reproduction of images can be done with ease, copy detection has increasingly become important. In the duplication process, image modifications are likely to occur and some alterations are deliberate and can be viewed as attacks. A wide range of copy detection techniques has been proposed. In our study, content-based copy detection, which basically applies DCT-based features for images, namely, pixel values, edges, texture information and frequency-domain component distribution, is employed. Experiments are carried out to evaluate robustness and sensitivity of DCT-based features from attacks. As different types of DCT-based features hold different pieces of information, how features and attacks are related can be shown in their robustness and sensitivity. Rather than searching for proper features, use of robustness and sensitivity is proposed here to realize how the attacked features have changed when an image attack occurs. The experiments show that, out of ten attacks, the neural networks are able to detect seven attacks namely, Gaussian noise, S&P noise, Gamma correction (high), blurring, resizing (big), compression and rotation with mostly related to their sensitive features.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.4
• /
• pp.315-320
• /
• 2015

This paper describes the implemenation of localization and visualization scheme to find out an ultrasonic source caused by defects of a power distribution line equipment. To increase the fault detection performance, $2{\times}4$ sensor array is configured with MEMS ultrasonic sensors, and from the sensor signals aquired, the azimuth and elevation angles of the ultrasonic source is estimated based on the delay-sum beam forming method. Also, to visualize the estimated location, it is marked on the background image. Experimental results show applicability of the present technique.

• Convergence Security Journal
• /
• v.13 no.1
• /
• pp.3-10
• /
• 2013

In this paper, reliability software model considering fault detection rate based on observations from the process of software product testing was studied. Adding new fault probability using the S-shaped distribution model that is widely used in the field of reliability problems presented. When correcting or modifying the software, finite failure non-homogeneous Poisson process model was used. In a software failure data analysis considering the time-dependent fault detection rate, the parameters estimation using maximum likelihood estimation of failure time data and reliability make out.

• Journal of Korean Neurosurgical Society
• /
• v.51 no.2
• /
• pp.120-123
• /
• 2012

Objective : Percutaneous vertebroplasty (PVP) is usually carried out under three-dimensional (2D) fluoroscopic guidance. However, operative complications or bone cement distribution might be difficult to assess on the basis of only 2D radiographic projection images. We evaluated the feasibility of performing an intraoperative and postoperative examination in patients undergoing PVP by using three-dimensional (3D) reconstructive C-arm. Methods : Standard PVP procedures were performed on 14 consecutive patients by using a Siremobil Iso-$C^{3D}$ and a multidetector computed tomography machine. Post-processing of acquired volumetric datasets included multiplanar reconstruction (MPR) and surface shaded display (SSD). We analyzed intraoperative and immediate postoperative evaluation of the needle trajectory and bone cement distribution. Results : The male : female ratio was 2 : 12; mean age of patients, 70 (range, 77-54) years; and mean T score, -3.4. The mean operation time was 52.14 min, but the time required to perform and post-process the rotational acquisitions was 7.76 min. The detection of bone cement distribution and leakage after PVP by using MPR and SSD was possible in all patients. However, detection of the safe trajectory for needle insertion was not possible. Conclusion : 3D rotational image acquisition can enable intra- or post-procedural assessment of vertebroplasty procedures for the detection of bone cement distribution and leakage. However, it is difficult to assess the safe trajectory for needle insertion.

• Proceedings of the KSME Conference
• /
• 2004.11a
• /
• pp.13-18
• /
• 2004

It is important to estimate the distribution of intensity of a magnetic field for application of magnetic method to industrial nondestructive evaluation. Magnetic camera provides the distribution of a quantitative magnetic field with homogeneous lift-off and same spatial resolution. Leakage magnetic flux near the crack on the specimen could be amplified by 3-dimensional magnetic fluid and zoom in and out of measurement area. This study introduces the experimental consideration of the effects of lens for concentrating of magnetic flux. The experimental results showed that the magnetic fluid has sufficient lens effect for magnetic camera and effect of improvement in probability of detection.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.33 no.12C
• /
• pp.1082-1087
• /
• 2008

Maximum Likelihood (ML) detection is well known to exhibit better bit-error-rate (BER) than many other detectors for multiple-input multiple-output (MIMO) channel. However, ML detection has been shown a difficult problem due to its NP-hard problem. It means that there is no known algorithm which can find the optimal solution in polynomial-time. In this paper, Semi-Definite relaxation (SDR) is iteratively applied to ML detection problem. The probability distribution can be obtained by survival eigenvector out of the dominant eigenvalue term of the optimal solution. The probability distribution which is yielded by SDR is recurred to the received signal. Our approach can reach to nearly ML performance.