• Journal of the Korean Society of Marine Environment & Safety
• /
• v.26 no.6
• /
• pp.601-605
• /
• 2020

In large ships with complex structures, it is difficult to locate workers. In particular, it is not easy to detect when a worker falls down, making it difficult to respond quickly. Thus, research is being conducted to detect fallen workers using a camera or by attaching a device to the body. Existing image-based fall detection systems have been designed to detect a person's body parts; hence, it is difficult to detect them in various ships and postures. In this study, the entire fall area was extracted and deep learning was used to detect the fallen shipworker based on the image. The data necessary for learning were obtained by recording falling states at the shipyard. The amount of learning data was augmented by flipping, resizing, and rotating the image. Performance evaluation was conducted with precision, reproducibility, accuracy, and a low error rate. The larger the amount of data, the better the precision. In the future, reinforcing various data is expected to improve the effectiveness of camera-based fall detection models, and thus improve safety.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.19 no.7
• /
• pp.7-15
• /
• 2020

Recently, the fourth industrial revolution has been presented as a new paradigm and additive manufacturing (AM) has become one of the most important topics. For this reason, process monitoring for each cross-sectional layer of additive metal manufacturing is important. Particularly, deep learning can train a machine to analyze, optimize, and repair defects. In this paper, image classification is proposed by learning images of defects in the metal cross sections using the convolution neural network (CNN) image labeling algorithm. Defects were classified into three categories: crack, porosity, and hole. To overcome a lack-of-data problem, the amount of learning data was augmented using a data augmentation algorithm. This augmentation algorithm can transform an image to 180 images, increasing the learning accuracy. The number of training and validation images was 25,920 (80 %) and 6,480 (20 %), respectively. An optimized case with a combination of fully connected layers, an optimizer, and a loss function, showed that the model accuracy was 99.7 % and had a success rate of 97.8 % for 180 test images. In conclusion, image labeling was successfully performed and it is expected to be applied to automated AM process inspection and repair systems in the future.

• Journal of Internet Computing and Services
• /
• v.20 no.6
• /
• pp.65-72
• /
• 2019

Data augmentation is a method that increases the amount of data through various algorithms based on a small amount of sample data. When machine learning and deep learning techniques are used to solve real-world problems, there is often a lack of data sets. The lack of data is at greater risk of underfitting and overfitting, in addition to the poor reflection of the characteristics of the set of data when learning a model. Thus, in this paper, through the layer-wise data augmenting method at each layer of deep neural network, the proposed method produces augmented data that is substantially meaningful and shows that the method presented by the paper through experimentation is effective in the learning of the model by measuring whether the method presented by the paper improves classification accuracy.

• The Journal of the Convergence on Culture Technology
• /
• v.4 no.1
• /
• pp.27-32
• /
• 2018

Recently, diverse bio-growth and object production experience learning contents have been developed for AR-based storytelling. In this paper, we will develop a storytelling-based growth type content using 3D printer which can add interest in and motivation of experiential learning. It is possible to increase the value of the characters produced by 3D printer as they can be used as personal collections. The bio-growth type content enables the learners and users to experience the local culture, history, and tourism ecology, which can enhance the educational effect by promoting their understanding in a fun way. The purpose of this study is to develop an application for experiential learning contents which creates a virtual environment in an object smart device where you can experience the culture and ecology of a learning community, and then produce the output using the 3D printer to keep as a personal souvenir. The developed contents application for experiential learning can be utilized not only for students but also for tourists. It is expected to serve as a source for further development of various contents through 3D printing technology.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.6
• /
• pp.578-586
• /
• 2021

We propose a semi-supervised learning method based on Variational AutoEncoder (VAE) and Unsupervised Data Augmentation (UDA) to improve the performance of an end-to-end speech recognizer. In the proposed method, first, the VAE-based augmentation model and the baseline end-to-end speech recognizer are trained using the original speech data. Then, the baseline end-to-end speech recognizer is trained again using data augmented from the learned augmentation model. Finally, the learned augmentation model and end-to-end speech recognizer are re-learned using the UDA-based semi-supervised learning method. As a result of the computer simulation, the augmentation model is shown to improve the Word Error Rate (WER) of the baseline end-to-end speech recognizer, and further improve its performance by combining it with the UDA-based learning method.

• Journal of Digital Convergence
• /
• v.14 no.5
• /
• pp.85-93
• /
• 2016

Convergent education is essential to develop consilient thinking skills, ability to recreate information and knowledge, and problem-solving skills which are demanded in society of convergent knowledge. Accordingly, this study is going to develop a convergent class model of augmented reality and art based on consilient knowledge. Teaching model is designed based on the ADDIE model, which was operated by opening a real class in order to verify the validity. The result showed a high satisfaction of learners showed the ability to adapt to the major areas associated with the cultivation of learners. Characteristics of augmented reality medium were found to enable learners to analyze a new phenomenon and to fuse the necessary knowledge by inducing them to actively interact by the their intention in learning. Therefore, it is possible to cultivate creative and convergent persons of ability equipped with more flexible and creative thinking ability and discernment through deepened education for recognizing and understanding convergent cases between art and scientific technology. There is a study on the validation and the convergence of different subjects in terms of a variety of aspects, left behind of this study.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.1
• /
• pp.355-360
• /
• 2023

Recently, with the development of digital, children are exposed to a lot of media media. Reading activity decreases, making it difficult to acquire information from books or organize and remember acquired information. Since education with augmented reality increases children's participation and immersion in learning, we proposed a reading activity application with augmented reality technology to increase children's reading interest and creativity. In addition, based on the five nurturing processes, we designed a play-oriented reading activity for the application. In addition to the application suggestions, we designed a play-centered lesson model so that it can be used in actual lessons. In order to analyze the conceptual thinking framework according to the lesson model design, we visited an actual daycare center and conducted a class attended by an expert. Experts who participated were asked to fill out a pre-produced questionnaire to review the suitability of the reading activity class model and operation, and the feasibility was examined. Our lesson model design was based on limited book content, and due to ethical concerns, large-scale experiments with children could not be conducted, so the results of the study are not representative of the whole. However, it is significant that the possibility of running a new reading activity class based on the Nuri course has been examined and proposed.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.11
• /
• pp.61-69
• /
• 2019

In this paper, we propose a multisensory(i.e., visual-auditory-tactile) based AR system for the education of cultural heritage. The proposed system provides a multisensory interaction by designing a user to experience with a 3D printed artifact which is mapped by a virtual 3D content of digital heritage. Compared with the existing systems of cultural heritage education based on augmented reality(AR) technology, this system focused on not only providing learning experience via a sense of visual and auditory, but also a sense of tactile. Furthermore, since this systems mainly provided the direct interactions using a 3D printed model, it gives a higher degree of realism than existing system that use touch or click motions on a 2D display of mobile phones and tablets. According to a result of user testing, we concluded that the proposed system delivered the excellent presence and learning flow to users. Particularly, from the usability evaluation, a 3D printed target artifact which is similar in shape to original heritage artifact, achieved the highest scores among the various tested targets.

• Informatization Policy
• /
• v.25 no.4
• /
• pp.3-36
• /
• 2018

This article discusses the possibilities and limitations of ICT education by reviewing the previous research on its various aspects including educational goals, contents, methods, and evaluation. First, when it comes to its educational goal, the prior studies suggest that ICT education aims to nurture digital citizenship among students and to enable them to participate in different sectors of our society. ICT education characterizes the core capacities of its future learners as 'lifelong learners,' 'information producers/consumers,' and 'local/global citizens.' Second, in regard to the educational content of ICT education, researchers investigate SW education importantly: They develop the educational programs and examine the effectiveness of those programs. However, to ensure the relevancy of the educational contents to the future society, institutional support is imperative including facilitating educators' capacities and synchronizing ICT education with subject education. Third, as the educational methods, various ICTs such as flipped learning and augmented reality (AR) are being applied to actual classroom teaching. Research on the educational methods, which is the most vibrant area in the ICT education scholarship, is expected to improve the previous educational methods and to lead the qualitative development of ICT education. Fourth, the previous discussion on the educational evaluation focuses on computer-based evaluations. Educational evaluation using ICT will enable educators to assess the characteristics and achievement of an individual learner accurately and to lead them to apply a teaching-learning process effectively, which will ultimately enhance the effectiveness of educational evaluation. Along with the overall review on the possibilities of ICT education, this article discusses the limitations of the current ICT education and its implications for educational inequalities.

• Journal of The Korean Association For Science Education
• /
• v.43 no.2
• /
• pp.181-190
• /
• 2023

In this study, we compared discourse by environments for using tools in terms of participation types, discourse types, and knowledge building processes. 24 first-year high school students were divided into six groups. They were assigned to the sharing tools environment, which used one marker and one smart device, or the individual tools environment, which used markers and smart devices individually. Students participated in small group learning using AR application based on the concept of chemical bonding. All classes were video- and audio-taped. Semi-structured interviews were conducted with six students who voluntarily agreed. The results of the study revealed that the sharing tools environment had a high proportion of one-student dominating type, while the individual tools environment had a high proportion of partly participating type and most students participating type. In the individual tools environment, the ratio of knowledge sharing and knowledge construction discourse was similar compared to the tool sharing environment, and the sub-discourse types were also diverse. In the sharing tools environment, only some students had a meaningful knowledge building process. On the other hand, in the individual tools environment, most of the group members constructed knowledge about the target concept, and had a meaningful knowledge building process. In addition, the misconceptions that appeared to some group members were corrected through small group discussions.