• International Journal of Advanced Culture Technology
• /
• v.7 no.4
• /
• pp.321-326
• /
• 2019

As AI technology is recently introduced into various fields, it is being applied to the fashion field. This paper proposes a system for recommending cody clothes suitable for a user's selected clothes. The proposed system consists of user app, cody recommendation module, and server interworking of each module and managing database data. Cody recommendation system classifies clothing images into 80 categories composed of feature combinations, selects multiple representative reference images for each category, and selects 3 full body cordy images for each representative reference image. Cody images of the representative reference image were determined by analyzing the user's preference using Google survey app. The proposed algorithm classifies categories the clothing image selected by the user into a category, recognizes the most similar image among the classification category reference images, and transmits the linked cody images to the user's app. The proposed system uses the ResNet-50 model to categorize the input image and measures similarity using ORB and HOG features to select a reference image in the category. We test the proposed algorithm in the Android app, and the result shows that the recommended system runs well.

• International Journal of Advanced Culture Technology
• /
• v.9 no.2
• /
• pp.86-90
• /
• 2021

Sentiment refers to a person's thoughts, opinions, and feelings toward an object. Sentiment analysis is a process of collecting opinions on a specific target and classifying them according to their emotions, and applies to opinion mining that analyzes product reviews and reviews on the web. Companies and users can grasp the opinions of public opinion and come up with a way to do so. Recently, natural language processing models using the Transformer structure have appeared, and Google's BERT is a representative example. Afterwards, various models came out by remodeling the BERT. Among them, the Facebook AI team unveiled the XLM-R (XLM-RoBERTa), an upgraded XLM model. XLM-R solved the data limitation and the curse of multilinguality by training XLM with 2TB or more refined CC (CommonCrawl), not Wikipedia data. This model showed that the multilingual model has similar performance to the single language model when it is trained by adjusting the size of the model and the data required for training. Therefore, in this paper, we study the improvement of Korean sentiment analysis performed using a pre-trained XLM-R model that solved curse of multilinguality and improved performance.

• Journal of Platform Technology
• /
• v.10 no.1
• /
• pp.20-28
• /
• 2022

Artificial intelligence-based health data verification has become an essential element not only to help clinical research, but also to develop new treatments. Since the US Food and Drug Administration (FDA) approved the marketing of medical devices that detect mild abnormal diabetic retinopathy in adult diabetic patients using artificial intelligence in the field of medical diagnosis, tests using artificial intelligence have been increasing. In this study, an artificial intelligence model based on image classification was created using a Teachable Machine supported by Google, and a predictive model was completed through learning. This not only facilitates the early detection of cataracts among eye diseases occurring among patients with chronic diseases, but also serves as basic research for developing a digital personal health healthcare app for eye disease prevention as a healthcare program for eye health.

• Journal of Distribution Science
• /
• v.21 no.8
• /
• pp.83-96
• /
• 2023

Purpose: This study examined the relationship between the factors (Credibility, Usability) and user Innovativeness of the ChatGPT on TRA (Theory of Reasoned Action; Subjective Norm, Attitude) and Switching Intention. TRA and Innovation Diffusion Theory (IDT) were used. Research design, data and methodology: From April 26 to 27, 2023, an online panel survey agency was commissioned to conduct a survey of GhatGPT users in their 20s and 40s in Korea, and a total of 210 people were used for the final analysis. Verification of the research model was performed using SPSS and AMOS. Results: First, ChatGPT factors (Credibility, Usability) were found to have positive effects on TRA (Subjective Norm, Attitude). Second, ChatGPT user Innovativeness was found to have a positive effect on TRA (Subjective Norm, Attitude). Third, ChatGPT users' TRA (Subjective Norm, Attitude) were found to have positive effects on Switching Intention. Conclusions: These results mean that the superior Usability and Credibility of ChatGPT and the Innovativeness of users have a significant effect on the Switching Intention from existing Portal Service (Naver, Google, Daum, etc.) to ChatGPT. Generative AI such as ChatGPT should strive to develop various services such as improving the convenience of functions so that innovative users can use them easily and conveniently in order to provide services that meet expectations.

• Fashion & Textile Research Journal
• /
• v.25 no.6
• /
• pp.739-744
• /
• 2023

With the popularization of 5G networks and the development of AI (artificial intelligence) technology, Metaverse, which creates production capacity by combining virtual space and reality, is attracting attention. In this study, we searched for makeup applications with more than 100 million downloads from October 11, 2020 to November 3, 2020 through the Google Play Store. As a result of the search, four applications were found: YouCam Makeup, YouCam Perfect, Beauty Plus, and Sweet Snap. Based on the functions provided by the four applications, we attempted to suggest makeup functions applicable to Zepeto's avatar. Functions for the eyes (eyeliner, eyelashes, mascara, eye shadow, eye shape, eyebrow shape, lenses, double eyelids), functions for the nose (nose shape), functions for the mouth (lipstick, lip shape, smile function) ) Functions corresponding to the facial contour (contour, skin foundation, blusher, shading, highlighter, face painting, theme makeup) and functions corresponding to the body (body adjustment) were proposed. This study is the first in the beauty field to propose a method of applying the functions of the Metaverse platform as the importance of digital platforms is highlighted, and is the first to propose a makeup function applied to the Metaverse so that it can be used as important basic data in the future.

• Phonetics and Speech Sciences
• /
• v.15 no.4
• /
• pp.71-80
• /
• 2023

This study primarily aimed to develop an automated stuttering identification and classification method using artificial intelligence technology. In particular, this study aimed to develop a deep learning-based identification model utilizing the convolutional neural networks (CNNs) algorithm for Korean speakers who stutter. To this aim, speech data were collected from 9 adults who stutter and 9 normally-fluent speakers. The data were automatically segmented at the phrasal level using Google Cloud speech-to-text (STT), and labels such as 'fluent', 'blockage', prolongation', and 'repetition' were assigned to them. Mel frequency cepstral coefficients (MFCCs) and the CNN-based classifier were also used for detecting and classifying each type of the stuttered disfluency. However, in the case of prolongation, five results were found and, therefore, excluded from the classifier model. Results showed that the accuracy of the CNN classifier was 0.96, and the F1-score for classification performance was as follows: 'fluent' 1.00, 'blockage' 0.67, and 'repetition' 0.74. Although the effectiveness of the automatic classification identifier was validated using CNNs to detect the stuttered disfluencies, the performance was found to be inadequate especially for the blockage and prolongation types. Consequently, the establishment of a big speech database for collecting data based on the types of stuttered disfluencies was identified as a necessary foundation for improving classification performance.

• Journal of Internet Computing and Services
• /
• v.21 no.3
• /
• pp.123-131
• /
• 2020

Along with the rapid development of artificial intelligence technology, natural language processing, which deals with human language, is also actively studied. In particular, BERT, a language model recently proposed by Google, has been performing well in many areas of natural language processing by providing pre-trained model using a large number of corpus. Although BERT supports multilingual model, we should use the pre-trained model using large amounts of Korean corpus because there are limitations when we apply the original pre-trained BERT model directly to Korean. Also, text contains not only vocabulary, grammar, but contextual meanings such as the relation between the front and the rear, and situation. In the existing natural language processing field, research has been conducted mainly on vocabulary or grammatical meaning. Accurate identification of contextual information embedded in text plays an important role in understanding context. Knowledge graphs, which are linked using the relationship of words, have the advantage of being able to learn context easily from computer. In this paper, we propose a system to extract Korean contextual information using pre-trained BERT model with Korean language corpus and knowledge graph. We build models that can extract person, relationship, emotion, space, and time information that is important in the text and validate the proposed system through experiments.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.3
• /
• pp.268-278
• /
• 2021

Technologies that are leading the fourth industrial revolution, such as the Internet of Things (IoT), big data, artificial intelligence (AI), and cyber-physical systems (CPS) are developing and generalizing. The demand to improve productivity, economy, safety, etc., is spreading in various industrial fields by applying these technologies. Digital twins are attracting attention as an important technology trend to meet demands and is one of the top 10 tasks of the Korean version of the New Deal. In this study, papers, magazines, reports, and other literature were searched using Google. In order to investigate the contribution or role of geospatial information in the digital twin application, the definition of a digital twin, we investigated technology trends of domestic and foreign companies; the components of digital twins required in manufacturing, plants, and smart cities; and the core techniques for driving a digital twin. In addition, the contributing contents of geospatial information were summarized by searching for a sentence or word linked between geospatial-related keywords (i.e., Geospatial Information, Geospatial data, Location, Map, and Geodata and Digital Twin). As a result of the survey, Geospatial information is not only providing a role as a medium connecting objects, things, people, processes, data, and products, but also providing reliable decision-making support, linkage fusion, location information provision, and frameworks. It was found that it can contribute to maximizing the value of utilization of digital twins.

• Journal of Intelligence and Information Systems
• /
• v.29 no.4
• /
• pp.165-184
• /
• 2023

Recently, Deep learning analysis of unstructured text data using language models, such as Google's BERT and OpenAI's GPT has shown remarkable results in various applications. Most language models are used to learn generalized linguistic information from pre-training data and then update their weights for downstream tasks through a fine-tuning process. However, some concerns have been raised that privacy may be violated in the process of using these language models, i.e., data privacy may be violated when data owner provides large amounts of data to the model owner to perform fine-tuning of the language model. Conversely, when the model owner discloses the entire model to the data owner, the structure and weights of the model are disclosed, which may violate the privacy of the model. The concept of offsite tuning has been recently proposed to perform fine-tuning of language models while protecting privacy in such situations. But the study has a limitation that it does not provide a concrete way to apply the proposed methodology to text classification models. In this study, we propose a concrete method to apply offsite tuning with an additional classifier to protect the privacy of the model and data when performing multi-classification fine-tuning on Korean documents. To evaluate the performance of the proposed methodology, we conducted experiments on about 200,000 Korean documents from five major fields, ICT, electrical, electronic, mechanical, and medical, provided by AIHub, and found that the proposed plug-in model outperforms the zero-shot model and the offsite model in terms of classification accuracy.

• International Journal of Advanced Culture Technology
• /
• v.8 no.3
• /
• pp.61-72
• /
• 2020

The fear of crime, discussed in the early 1960s in the United States, is a psychological response, such as anxiety or concern about crime, the potential victim of a crime. These anxiety factors lead to the burden of the individual in securing the psychological stability and indirect costs of the crime against the society. Fear of crime is not a good thing, and it is a part that needs to be adjusted so that it cannot be exaggerated and distorted by the policy together with the crime coping and resolution. This is because fear of crime has as much harm as damage caused by criminal act. Eric Pawson has argued that the popular impression of violent crime is not formed because of media reports, but by official statistics. Therefore, the police should watch and analyze news related to fear of crime to reduce the social cost of fear of crime and prepare a preemptive response policy before the people have 'fear of crime'. In this paper, we propose a deep - based news classification system that helps police cope with crimes related to crimes reported in the media efficiently and quickly and precisely. The goal is to establish a system that can quickly identify changes in security issues that are rapidly increasing by categorizing news related to crime among news articles. To construct the system, crime data was learned so that news could be classified according to the type of crime. Deep learning was applied by using Google tensor flow. In the future, it is necessary to continue research on the importance of keyword according to early detection of issues that are rapidly increasing by crime type and the power of the press, and it is also necessary to constantly supplement crime related corpus.