• The Journal of Korea Robotics Society
• /
• v.14 no.2
• /
• pp.139-149
• /
• 2019

In this paper, we present auto-annotation tool and synthetic dataset using 3D CAD model for deep learning based object detection. To be used as training data for deep learning methods, class, segmentation, bounding-box, contour, and pose annotations of the object are needed. We propose an automated annotation tool and synthetic image generation. Our resulting synthetic dataset reflects occlusion between objects and applicable for both underwater and in-air environments. To verify our synthetic dataset, we use MASK R-CNN as a state-of-the-art method among object detection model using deep learning. For experiment, we make the experimental environment reflecting the actual underwater environment. We show that object detection model trained via our dataset show significantly accurate results and robustness for the underwater environment. Lastly, we verify that our synthetic dataset is suitable for deep learning model for the underwater environments.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2020.07a
• /
• pp.329-330
• /
• 2020

본 논문에서는 Azure 클라우드 플랫폼의 가상서버 호스팅을 이용해 데이터 수집 환경을 구축하고, Azure에서 제공하는 자동화된 기계학습(Automated Machine Learning, AutoML)을 기반으로 데이터 분석 방법에 관한 연구를 수행했다. 가상 서버 호스팅 환경에 LAMP(Linux, Apache, MySQL, PHP)를 설치하여 데이터 수집환경을 구축했으며, 수집된 데이터를 Azure AutoML에 적용하여 자동화된 기계학습을 수행했다. Azure AutoML은 소모적이고 반복적인 기계학습 모델 개발을 자동화하는 프로세스로써 기계학습 솔루션 구현하는데 시간과 자원(Resource)를 절약할 수 있다. 특히, AutoML은 수집된 데이터를 분류와 회귀 및 예측하는데 있어서 학습점수(Training Score)를 기반으로 보유한 데이터에 가장 적합한 기계학습 모델의 순위를 제공한다. 이는 데이터 분석에 필요한 기계학습 모델을 개발하는데 있어서 개발 초기 단계부터 코드를 설계하지 않아도 되며, 전체 기계학습 시스템을 개발 및 구현하기 전에 모델의 구성과 시스템을 설계해볼 수 있기 때문에 매우 효율적으로 활용될 수 있다. 본 논문에서는 NPU(Neural Processing Unit) 학습에 필요한 데이터 수집 환경에 관한 연구를 수행했으며, Azure AutoML을 기반으로 데이터 분류와 회귀 등 가장 효율적인 알고리즘 선정에 관한 연구를 수행했다.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.728-735
• /
• 2022

Roofers get exposed to increased risk of knee musculoskeletal disorders (MSDs) at different phases of a sloped shingle installation task. As different phases are associated with different risk levels, this study explored the application of machine learning for automated classification of seven phases in a shingle installation task using knee kinematics and roof slope information. An optical motion capture system was used to collect knee kinematics data from nine subjects who mimicked shingle installation on a slope-adjustable wooden platform. Four features were used in building a phase classification model. They were three knee joint rotation angles (i.e., flexion, abduction-adduction, and internal-external rotation) of the subjects, and the roof slope at which they operated. Three ensemble machine learning algorithms (i.e., random forests, decision trees, and k-nearest neighbors) were used for training and prediction. The simulations indicate that the k-nearest neighbor classifier provided the best performance, with an overall accuracy of 92.62%, demonstrating the considerable potential of machine learning methods in detecting shingle installation phases from workers knee joint rotation and roof slope information. This knowledge, with further investigation, may facilitate knee MSD risk identification among roofers and intervention development.

• The Korean Journal of Emergency Medical Services
• /
• v.21 no.2
• /
• pp.39-50
• /
• 2017

Purpose: This study measured awareness of automated external defibrillators (AED) among students majoring in health-related versus non-health-related fields. Methods: A total of 577 students filled out a questionnaire on awareness of AEDs between June 12 and June 28, 2017. Using SPSS 23.0, data were analyzed using descriptive statistics, chi-square tests, and t-tests. Results: In response to a question about awareness of AED, 25.0% of students majoring in a health-related and 38.6% of those majoring in a non-health-related field answered on "I have seen or heard" and "I have no idea". In response to a question on perception on AED use, 82.4% of students majoring in a health-related field and 88.1% of those in a non-health-related field reported that they thought perception of AED use was not universal. In terms of experience with education on the use of AED, 30.2% of health-related majors and 49.7% of non-health-related majors had not received any training on the use of AED. The average overall score regarding knowledge about AED was 8.69 for health-related majors, compared to 7.79 for non-health-related majors. Conclusion: In order to improve awareness regarding AED use, education on importance and necessity of AED should be emphasized and implemented consistently by the mass media.

• The korean journal of orthodontics
• /
• v.51 no.2
• /
• pp.77-85
• /
• 2021

Objective: To evaluate the accuracy of a multi-stage convolutional neural network (CNN) model-based automated identification system for posteroanterior (PA) cephalometric landmarks. Methods: The multi-stage CNN model was implemented with a personal computer. A total of 430 PA-cephalograms synthesized from cone-beam computed tomography scans (CBCT-PA) were selected as samples. Twenty-three landmarks used for Tweemac analysis were manually identified on all CBCT-PA images by a single examiner. Intra-examiner reproducibility was confirmed by repeating the identification on 85 randomly selected images, which were subsequently set as test data, with a two-week interval before training. For initial learning stage of the multi-stage CNN model, the data from 345 of 430 CBCT-PA images were used, after which the multi-stage CNN model was tested with previous 85 images. The first manual identification on these 85 images was set as a truth ground. The mean radial error (MRE) and successful detection rate (SDR) were calculated to evaluate the errors in manual identification and artificial intelligence (AI) prediction. Results: The AI showed an average MRE of 2.23 ± 2.02 mm with an SDR of 60.88% for errors of 2 mm or lower. However, in a comparison of the repetitive task, the AI predicted landmarks at the same position, while the MRE for the repeated manual identification was 1.31 ± 0.94 mm. Conclusions: Automated identification for CBCT-synthesized PA cephalometric landmarks did not sufficiently achieve the clinically favorable error range of less than 2 mm. However, AI landmark identification on PA cephalograms showed better consistency than manual identification.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.304-311
• /
• 2022

Construction inspection is a crucial stage that ensures that all contractual requirements of a construction project are verified. The construction inspection capabilities among state highway agencies have been greatly affected due to budget reduction. As a result, efficient inspection practices such as risk-based inspection are required to optimize the use of limited resources without compromising inspection quality. Automated prioritization of textual requirements according to their criticality would be extremely helpful since contractual requirements are typically presented in an unstructured natural language in voluminous text documents. The current study introduces a novel model for predicting the risk level of requirements using machine learning (ML) algorithms. The ML algorithms tested in this study included naïve Bayes, support vector machines, logistic regression, and random forest. The training data includes sequences of requirement texts which were labeled with risk levels (such as very low, low, medium, high, very high) using the fuzzy logic systems. The fuzzy model treats the three risk factors (severity, probability, detectability) as fuzzy input variables, and implements the fuzzy inference rules to determine the labels of requirements. The performance of the model was examined on labeled dataset created by fuzzy inference rules and three different membership functions. The developed requirement risk prediction model yielded a precision, recall, and f-score of 78.18%, 77.75%, and 75.82%, respectively. The proposed model is expected to provide construction inspectors with a means for the automated prioritization of voluminous requirements by their importance, thus help to maximize the effectiveness of inspection activities under resource constraints.

• Dementia and Neurocognitive Disorders
• /
• v.21 no.4
• /
• pp.117-125
• /
• 2022

Background and Purpose: Medication adherence is essential for effective medical treatment. However, it is challenging for cognitively impaired patients. We investigated whether an automated telephone reminder service improves medication adherence and reduces the decline of cognitive function in isolated patients with cognitive impairment. Methods: This was a single-center, randomized clinical trial. We enrolled mild cognitive impairment (MCI) or Alzheimer's disease (AD) patients who lived alone or with a cognitively impaired spouse. We provided an automated telephone reminder service for taking medication to the intervention group for 6 months. The control group was provided with general guidelines for taking the medication every month. The participants underwent neuropsychological assessment at the beginning and end of the study. Statistical significance was tested using nonparametric Wilcoxon rank sum and Wilcoxon matched-pairs signed-rank tests. Results: Thirty participants were allocated randomly to groups, and data for 29 participants were analyzed. The mean age was 79.6 (standard deviation, 6.0) years and 79.3% of the participants were female. There was no significant difference in medication adherence between the 2 groups. However, a subgroup analysis among participants with more than 70% response rates showed better medication adherence compared to the control group (intervention: 94.6%; control: 90.2%, p=0.0478). There was no significant difference in the change in cognitive function between the 2 groups. Conclusions: If a patient's compliance is good, telephone reminders might be effective in improving medication adherence. It is necessary to develop reminder tools that can improve compliance for cognitively impaired patients.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.53 no.11
• /
• pp.596-603
• /
• 2004

This paper proposes an expert system based on the pattern recognition method which can enhance the accuracy and effectiveness of real-time bus reconfiguration strategy for the transfer of faulted load when a main transformer fault occurs in the automated substation. The minimum distance classification method is adopted as the pattern recognition method of expert system. The training pattern set is designed MTr by MTr to minimize the searching time for target load pattern which is similar to the real-time load pattern. But the control pattern set, which is required to determine the corresponding bus reconfiguration strategy to these trained load pattern set is designed as one table by considering the efficiency of knowledge base design because its size is small. The training load pattern generator based on load level and the training load pattern generator based on load profile are designed, which are can reduce the size of each training pattern set from max L/sup (m+f)/ to the size of effective level. Here, L is the number of load level, m and f are the number of main transformers and the number of feeders. The one reduces the number of trained load pattern by setting the sawmiller patterns to a same pattern, the other reduces by considering only load pattern while the given period. And control pattern generator based on exhaustive search method with breadth-limit is designed, which generates the corresponding bus reconfiguration strategy to these trained load pattern set. The inference engine of the expert system and the substation database and knowledge base is implemented in MFC function of Visual C++ Finally, the performance and effectiveness of the proposed expert system is verified by comparing the best-first search solution and pattern recognition solution based on diversity event simulations for typical distribution substation.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.39 no.5
• /
• pp.631-636
• /
• 2019

Over the last decade, many researchers have investigated a number of vision-based construction object detection algorithms for the purpose of construction site monitoring. However, previous methods require the ground truth labeling, which is a process of manually marking types and locations of target objects from training image data, and thus a large amount of time and effort is being wasted. To address this drawback, this paper proposes a vision-based construction object detection framework that employs an active learning technique while reducing manual labeling efforts. For the validation, the research team performed experiments using an open construction benchmark dataset. The results showed that the method was able to successfully detect construction objects that have various visual characteristics, and also indicated that it is possible to develop the high performance of an object detection model using smaller amount of training data and less iterative training steps compared to the previous approaches. The findings of this study can be used to reduce the manual labeling processes and minimize the time and costs required to build a training database.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.3
• /
• pp.133-139
• /
• 2021

Currently, the process of constructing a high-definition road map has a high proportion of manual labor, so there are limitations in construction time and cost. Research to automate map production with high-definition road maps using artificial intelligence is being actively conducted, but since the construction of training data for the map construction is also done manually, there is a need to automatically build training data. Therefore, in this study, after converting to images using point clouds acquired by a mobile mapping system, the road marking areas were extracted through image reclassification and overlap analysis using thresholds. Then, a methodology was proposed to automatically construct training data for deep learning data for the high-definition road map through the classification of the polygon types in the extracted regions. As a result of training 2,764 lane data constructed through the proposed methodology on a deep learning-based PointNet model, the training accuracy was 99.977%, and as a result of predicting the lanes of three color types using the trained model, the accuracy was 99.566%. Therefore, it was found that the methodology proposed in this study can efficiently produce training data for high-definition road maps, and it is believed that the map production process of road markings can also be automated.