Recently, several deep learning studies have been reported to automatically identify the location of diagnostic devices using endoscopic data. In previous studies, there was no design to determine whether the configuration of the dataset resulted in differences in the accuracy in which artificial intelligence models perform image classification. Studies that are based on large amounts of data are likely to have different results depending on the composition of the dataset or its proportion. In this study, we intended to determine the existence and extent of accuracy according to the composition of the dataset by compiling it into three main types using larynx, esophagus, gastroscopy, and laryngeal endoscopy images.

This research is to raise awareness of food safety by designing and supporting a hazard food information notification platform for consumers. To this end, the design was carried out by dividing the process into a data extraction process, an application screen design process, and a CNN-based food inference process. Data was collected through public data APIs and crawling, and it was sent to each activity screen designed for Android studios so that it could be output. As a result, when the platform is executed, information on hazardous food names, registration dates, food classification, manufacturing dates, recovery grades, recovery reasons, recovery methods, company names, barcode numbers, and packaging units can be intuitively and conveniently checked. In addition, CNN-based food inference processes allowed mobile cameras to infer harmful food and applied various quantization techniques such as Dynamic Range, Integer, and Float16 to compare the degree of improvement in inference performance. As a result, the group that applied basic quantization and treated device resources with GPU showed the greatest improvement in inference performance. Through this platform, it is expected that the reliability of food safety will be improved by making it more convenient for consumers to recognize food risks.

Currently, fine dust forecast models are operated by dividing the country into 19 regions. Therefore, it is important to reduce the learning time and the number of models as well as accurate forecast performance to operate lots of forecast models. In this paper, we develop a categorical wide area model that outputs forecast results categorically and integrates the regions with similar regional characteristics. The proposed model improved the convergence rate by 223 times compared to the existing model, which outputs at a single concentration value, and reduced the number of forecast models by a third.

This study explored at the potentials of exhibition appreciation classes using the metaverse platform, and identified what types of interactions could occur when elementary students participate in the exhibition appreciation education using metaverse platform. We had a user observation research and FGI through implementation of actual elementary class and we analyzed problematic experiences from the viewpoint of user to content and user to user interaction. As a result, fifteen UX improvements were derived by Delphi technique for solving the problematic experiences and improving learner's interactivity.

This paper is a study on context-sensitive spelling error correction and uses the Korean WordNet (KorLex)[1] that defines the relationship between words as a graph to improve the performance of the correction[2] based on the vector information of the word embedded in the correction technique. The Korean WordNet replaced WordNet[3] developed at Princeton University in the United States and was additionally constructed for Korean. In order to learn a semantic network in graph form or to use it for learned vector information, it is necessary to transform it into a vector form by embedding learning. For transformation, we list the nodes (limited number) in a line format like a sentence in a graph in the form of a network before the training input. One of the learning techniques that use this strategy is Deepwalk[4]. DeepWalk is used to learn graphs between words in the Korean WordNet. The graph embedding information is used in concatenation with the word vector information of the learned language model for correction, and the final correction word is determined by the cosine distance value between the vectors. In this paper, In order to test whether the information of graph embedding affects the improvement of the performance of context- sensitive spelling error correction, a confused word pair was constructed and tested from the perspective of Word Sense Disambiguation(WSD). In the experimental results, the average correction performance of all confused word pairs was improved by 2.24% compared to the baseline correction performance.

With the development of media technology, virtual reality (VR) technology is widely used in education, medical care, aerospace, entertainment and other fields. Among them, application in teaching courseware is a relatively new topic. Compared with traditional coursewares, virtual games visualized and extruded abstract teaching contents. Thus it strengthened teaching effects and expanded dimensions of learning. We hypothesized that virtual coursewares could increase users'sense of presence and enhance their focus. In this study, virtual courseswares were compared with traditional coursewares. At the same time, its feasibility and advantages of application were analyzed through literature researching, practical researching and statistical analysis from questionnaires. Furthermore, we designed a teaching system for VR coursewares and explored its performance in multidimensional and contextual teaching situations. It was found that Virtual coursewares have changed the boring traditional teaching methods. The teaching content was displayed in the form of three-dimensional images, videos and sounds through VR equipment, which effectively improved teaching efficiency. In addition, the feasibility of virtual courseware was demonstrated through factor analysis in questionnaires. Compared with traditional teaching courseware, VR coursewares can attract students' attention and improve learning efficiency. It provides a good example and is valuable for the research of virtual realities in education.

This paper confirmed the technical reliability of mobile-based sarcopenia prediction and monitoring system. In implementing the developed system, we designed using only sensors built into a smartphone without a separate external device. The prediction system predicts the possibility of sarcopenia without visiting a hospital by performing the SARC-F survey, the 5-time chair stand test, and the rapid tapping test. The Monitoring system tracks and analyzes the average walking speed in daily life to quickly detect the risk of sarcopenia. Through this, it is possible to rapid detection of undiagnosed risk of undiagnosed sarcopenia and initiate appropriate medical treatment. Through prediction and monitoring system, the user may predict and manage sarcopenia, and the developed system can have a positive effect on reducing medical demand and reducing medical costs. In addition, collected data is useful for the patient-doctor communication. Furthermore, the collected data can be used for learning data of artificial intelligence, contributing to medical artificial intelligence and e-health industry.

With the development of the digital age, the moving poster shows the possibility of development and update. In this study, by analyzing the characteristics and types of moving posters, we studied the elements of motion graphic movement of moving posters. Through this specific case, we analyzed the characteristics of moving posters expressed in Memphis style to explore their strengths and applicability. In the current situation where the popularity of moving posters is increasing, it is good to develop and utilize the study of Memphis moving posters.

The purpose of this study is to help preschool children in the interface design of educational game APP at the psychological level. The conceptual model is constructed through the Technology Acceptance Model (TAM) to explore the perceived usefulness and perceived ease of use when children are playing educational game APP. The empirical study was conducted in the form of questionnaires collected after children used game. The research subjects were 3-6 years old children, and the research tool was the game app. Data collection under the guidance of their parents. The research shows that children educational game experience elements can effectively increase preschool children's psychological perception of educational games. Perceived ease of use also increases trust in educational games for children. The results were as follows: 1. Cronbach's Alpha and KMO were 0.969 and 0.955, and the P value was significant, which passed the reliability and validity test. 2. Through confirmatory factor analysis (Model fit index, Composite reliability, discriminant validity), we found that user experience is closely related to perceived usefulness and perceived ease of use. 3. The path analysis of the relationship proves that perceived ease of use play a key role in trusting preschool children educational game APP.

Emotional experience (James, 1884)[1] can affect people's behavior. There are few types of research on VR marketing(Maojun Zhou, Zeru Yan, 2018)[2] design based on emotional experience. This article is based on emotional evaluation theory and empirical research, and the VR marketing case "Buy+" online shopping platform (Wu Yongyi, 2016). It is concluded that there are three levels of emotional experience definition on VR marketing which decompose the features of the VR marketing design of "Buy+ as an online-shop" correspondingly and find out the design features of VR marketing from the perspective of emotional experience. Finally, through the analysis of the questionnaire data, it verified that vividness, functionality and effectiveness could represent the features of VR marketing design. Moreover, it analyzed the correlation among these factors. Vividness and functionality have the closest relationship among them. The definition, the components, and the correlation of the three-layer emotional experience obtained from this research can provide theoretical support and reference for other VR marketing designs.

Recently, the demand to interpret image data with artificial intelligence in various fields is rapidly increasing. Object recognition and detection techniques using deep learning are mainly used, and video integration analysis to determine unstructured object recognition is a particularly important problem. In the case of natural disasters or social disasters, there is a limit to the object recognition structure alone because it has an unstructured shape. In this paper, we propose intelligent video integration analysis system that can recognize unstructured objects based on video turning point and object detection. We also introduce a method to apply and evaluate object recognition using virtual augmented images from 2D to 3D through GAN.