• Annual Conference of KIPS
• /
• 2008.11a
• /
• pp.112-113
• /
• 2008

유방암은 유럽과 미국을 비롯해 전 세계적으로 증가하고 있으며 최근 우리나라에서도 가장 급속하게 늘고 있는 여성암중에 하나이다. 본 연구에서는 먼저 grey level co-occurrence matrix(GLCM)을 적용하여 유방영역을 분할한 후, median filter 를 적용하여 잡음을 제거하였다. 전처리 수행 후, 2차미분 행렬을 이용할여 미소석회화 부분을 강조한 후, 가우시안 정규분포도를 적용하여 미소석회화 후보군을 검출하였다. 검출된 후보군은 8 개의 feature 들을 적용하여 미소석회화를 최종 결정하였다. 본 연구를 통해서 조기 유방암 진단을 위한 발전된 미소석회화 검출 방법을 제안하였다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.45-46
• /
• 2024

본 논문에서는 Self-Attention 기반 딥러닝 기법인 Dense Prediction Transformer(DPT) 모델을 MVTec Anomaly Detection(MVTec AD) 데이터셋에 적용하여 실제 산업 제품 이미지 내 이상 부분을 분할하는 연구를 진행하였다. DPT 모델의 적용을 통해 기존 Convolutional Neural Network(CNN) 기반 이상 탐지기법의 한계점인 지역적 Feature 추출 및 고정된 수용영역으로 인한 문제를 개선하였으며, 실제 산업 제품 데이터에서의 이상 분할 시 기존 주력 기법인 U-Net의 구조를 적용한 최고 성능의 모델보다 1.14%만큼의 성능 향상을 보임에 따라 Self-Attention 기반 딥러닝 기법의 적용이 산업 제품 이상 분할에 효과적임을 입증하였다.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.2
• /
• pp.135-140
• /
• 2009

In this paper, we synthesized active sonar target signals based on highlights model, and then carried out target classification using the synthesized signals. If the target aspect angle is changed, the different signals are synthesized. To know the result, two different experiments are done. First, The classification results with respect to each aspect angle are shown. Second, the results in two group in aspect angle are acquired. Time domain feature extraction is done using matched filter and envelope detection. It shows the pattern of each highlights. Artificial neural networks and multi-class SVM are used for classifying target signals.

• Journal of Sensor Science and Technology
• /
• v.33 no.3
• /
• pp.139-146
• /
• 2024

Surface-enhanced Raman scattering (SERS) enables the detection of various types of π-conjugated biological and chemical molecules owing to its exceptional sensitivity in obtaining unique spectra, offering nondestructive classification capabilities for target analytes. Herein, we demonstrate an innovative strategy that provides significant machine learning (ML)-enabled predictive SERS platforms through surface-engineered graphene via complementary hybridization with Au nanoparticles (NPs). The hybridized Au NPs/graphene SERS platforms showed exceptional sensitivity (10^-7 M) due to the collaborative strong correlation between the localized electromagnetic effect and the enhanced chemical bonding reactivity. The chemical and physical properties of the demonstrated SERS platform were systematically investigated using microscopy and spectroscopic analysis. Furthermore, an innovative strategy employing ML is proposed to predict various analytes based on a featured Raman spectral database. Using a customized data-preprocessing algorithm, the feature data for ML were extracted from the Raman peak characteristic information, such as intensity, position, and width, from the SERS spectrum data. Additionally, sophisticated evaluations of various types of ML classification models were conducted using k-fold cross-validation (k = 5), showing 99% prediction accuracy.

• Journal of Sasang Constitutional Medicine
• /
• v.36 no.2
• /
• pp.39-48
• /
• 2024

Objectives The aim was to develop a method for detecting Sasang constitution based on the ratio of facial landmarks and provide an objective and reliable tool for Sasang constitution classification. Methods Facial images, KS-15 scores, and certainty scores were collected from subjects identified by Korean Medicine Data Center. Facial ratio landmarks were detected, yielding 2279 facial ratio features. Tree-based models were trained to classify Sasang constitution, and Shapley Additive Explanations (SHAP) analysis was employed to identify important facial features. Additionally, Body Mass Index (BMI) and personality questionnaire were incorporated as supplementary information to enhance model performance. Results Using the Tree-based models, the accuracy for classifying Taeeum, Soeum, and Soyang constitutions was 81.90%, 90.49%, and 81.90% respectively. SHAP analysis revealed important facial features, while the inclusion of BMI and personality questionnaire improved model performance. This demonstrates that facial ratio-based Sasang constitution analysis yields effective and accurate classification results. Conclusions Facial ratio-based Sasang constitution analysis provides rapid and objective results compared to traditional methods. This approach holds promise for enhancing personalized medicine in Korean traditional medicine.

• Journal of Astronomy and Space Sciences
• /
• v.41 no.3
• /
• pp.149-158
• /
• 2024

This study developed a machine learning-based methodology to classify gravitational wave (GW) signals from black hol-eneutron star (BH-NS) mergers by combining convolutional neural network (CNN) with conditional information for feature extraction. The model was trained and validated on a dataset of simulated GW signals injected to Gaussian noise to mimic real world signals. We considered all three types of merger: binary black hole (BBH), binary neutron star (BNS) and neutron starblack hole (NSBH). We achieved up to 96% correct classification of GW signals sources. Incorporating our novel conditional information approach improved classification accuracy by 10% compared to standard time series training. Additionally, to show the effectiveness of our method, we tested the model with real GW data from the Gravitational Wave Transient Catalog (GWTC-3) and successfully classified ~90% of signals. These results are an important step towards low-latency real-time GW detection.

• Journal of Wind Energy
• /
• v.13 no.3
• /
• pp.33-42
• /
• 2022

In this paper, a semi-supervised machine learning technique applied to actual field vibration data acquired from Jeju-do wind turbines for predictive diagnosis of abnormal conditions of offshore wind turbines is introduced. Semi-supervised machine learning, which combines un-supervised learning with supervised learning, can be used to perform anomaly detection in situations where sufficient fault data cannot be obtained. The signal processing results using the spectrogram of the original signal were shown, and external data were used to overcome the problem that disturbance reactions easily occurred due to the imbalance between the number of normal and abnormal data. Out of distribution (OOD), which uses external data, is a technology that is regarded as abnormal data that is unlikely to occur in reality, but we were able to use it by expanding it. By rearranging the distribution of data in this way, classification can be performed more robustly. Specifically, by observing the trends of the abnormal score and the change in the feature of the representation layer, continuous learning was performed through a mixture of existing and new data.

• Structural Monitoring and Maintenance
• /
• v.11 no.3
• /
• pp.203-217
• /
• 2024

Electro-mechanical impedance (EMI) technique is a low-cost structural damage detection method. It reflects structural damage through the change in admittance signal which contains the structural mechanical impedance information. The ambient temperature greatly affects the admittance signal, which hides the changes caused by structural damage and reduces the accuracy of damage identification. This study introduces a convolutional neural network to compensate for the temperature effect. The proposed method uses a framework that consists of a feature extraction network and a decoding network, and the original admittance signal with temperature information is used as the input. The output admittance signal is eliminated from the temperature effect, improving damage identification robustness. The admittance data simulated by the finite element model of the spatial grid structure is used to verify the effectiveness of the proposed method. The results show that the proposed method has advantages in identification accuracy compared with the damage index minimization method and the principal component analysis method.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.45 no.1
• /
• pp.25-36
• /
• 2008

Eye localization means localization of the center of the pupils, and is necessary for face recognition and related applications. Most of eye localization methods reported so far still need to be improved about robustness as well as precision for successful applications. In this paper, we propose a robust eye localization method using multi-scale Gabor feature vectors without big computational burden. The eye localization method using Gabor feature vectors is already employed in fuck as EBGM, but the method employed in EBGM is known not to be robust with respect to initial values, illumination, and pose, and may need extensive search range for achieving the required performance, which may cause big computational burden. The proposed method utilizes multi-scale approach. The proposed method first tries to localize eyes in the lower resolution face image by utilizing Gabor Jet similarity between Gabor feature vector at an estimated initial eye coordinates and the Gabor feature vectors in the eye model of the corresponding scale. Then the method localizes eyes in the next scale resolution face image in the same way but with initial eye points estimated from the eye coordinates localized in the lower resolution images. After repeating this process in the same way recursively, the proposed method funally localizes eyes in the original resolution face image. Also, the proposed method provides an effective illumination normalization to make the proposed multi-scale approach more robust to illumination, and additionally applies the illumination normalization technique in the preprocessing stage of the multi-scale approach so that the proposed method enhances the eye detection success rate. Experiment results verify that the proposed eye localization method improves the precision rate without causing big computational overhead compared to other eye localization methods reported in the previous researches and is robust to the variation of post: and illumination.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.37 no.3
• /
• pp.199-208
• /
• 2019

As the number of available satellites increases and technology advances, image information outputs are becoming increasingly diverse and a large amount of data is accumulating. In this study, we propose a change detection method for high-resolution satellite images that uses transfer learning and a deep learning network to overcome the limit caused by insufficient training data via the use of pre-trained information. The deep learning network used in this study comprises convolutional layers to extract the spatial and spectral information and convolutional long-short term memory layers to analyze the time series information. To use the learned information, the two initial convolutional layers of the change detection network are designed to use learned values from 40,000 patches of the ISPRS (International Society for Photogrammertry and Remote Sensing) dataset as initial values. In addition, 2D (2-Dimensional) and 3D (3-dimensional) kernels were used to find the optimized structure for the high-resolution satellite images. The experimental results for the KOMPSAT-3A (KOrean Multi-Purpose SATllite-3A) satellite images show that this change detection method can effectively extract changed/unchanged pixels but is less sensitive to changes due to shadow and relief displacements. In addition, the change detection accuracy of two sites was improved by using 3D kernels. This is because a 3D kernel can consider not only the spatial information but also the spectral information. This study indicates that we can effectively detect changes in high-resolution satellite images using the constructed image information and deep learning network. In future work, a pre-trained change detection network will be applied to newly obtained images to extend the scope of the application.