• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.31 no.6
• /
• pp.1215-1225
• /
• 2021

Although Artificial Intelligence (AI) techniques have shown impressive performance in various fields, they are vulnerable to adversarial examples which induce misclassification by adding human-imperceptible perturbations to the input. Previous studies to defend the adversarial examples can be classified into three categories: (1) model retraining methods; (2) input transformation methods; and (3) adversarial examples detection methods. However, even though the defense methods against adversarial examples have constantly been proposed, there is no research to classify the type of adversarial attack. In this paper, we proposed an adversarial attack family classification method based on dimensionality reduction and clustering. Specifically, after extracting adversarial perturbation from adversarial example, we performed Linear Discriminant Analysis (LDA) to reduce the dimensionality of adversarial perturbation and performed K-means algorithm to classify the type of adversarial attack family. From the experimental results using MNIST dataset and CIFAR-10 dataset, we show that the proposed method can efficiently classify five tyeps of adversarial attack(FGSM, BIM, PGD, DeepFool, C&W). We also show that the proposed method provides good classification performance even in a situation where the legitimate input to the adversarial example is unknown.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.175-201
• /
• 2023

In the midst of going through a non-face-to-face society, most of human activities narrowed down to the platform, restrictions on external activities are bringing the internal scalability of digital technology. Metaverse is virtually shifting reality and increasing the possibility of utilization in various areas. However, researches linked to the educational effects of metaverse, especially students with disabilities, are still an unknown area that lacks exploration. This paper focuses on the fact that metaverse-education is widening educational fields that meets the various needs of disabled students to realize social good and inclusive education, and communication effects such as resolving barriers to interaction are prominent. As a research method, examining literature research papers linked to AR/VR, metaverse with communication skills, interviews, articles, and columns by experts, and policy suggestions and implications for the special education was conducted. Although the limitations of research are confirmed, significant results are found on inclusive education, which provides educational maximizing effects and realizing human rights through direct immersive experience reflecting the Cone of Experience Theory. Hopefully follow-up studies on meta-edu for disabled students will be carried out in the future, and various interdisciplinary discussions are needed to carefully observe inclusive policies and benefits so that the socially vulnerable are not excluded from technologies in ICT society.

• Journal of Intelligence and Information Systems
• /
• v.28 no.1
• /
• pp.243-262
• /
• 2022

In this study, eye tracking was used for the appearance of the robot during the social robot design study. During the research, each part of the social robot was designated as AOI (Areas of Interests), and the user's attitude was measured through a design evaluation questionnaire to construct a design research model of the social robot. The data used in this study are Fixation, First Visit, Total Viewed, and Revisits as eye tracking indicators, and AOI (Areas of Interests) was designed with the face, eyes, lips, and body of the social robot. And as design evaluation questionnaire questions, consumer beliefs such as Face-highlighted, Human-like, and Expressive of social robots were collected and as a dependent variable was attitude toward robots. Through this, we tried to discover the mechanism that specifically forms the user's attitude toward the robot, and to discover specific insights that can be referenced when designing the robot.

• Information Systems Review
• /
• v.18 no.4
• /
• pp.69-88
• /
• 2016

Nowadays, organizations have adopted Smart Work to efficiently manage tasks, such as electronic document approval, customer management, and site inspection, without spatial-temporal constraints. Smartphones, which are commonly used in Smart Work, enable individuals to perform their jobs anytime and anywhere, thus blurring the boundary between work and non-work. To solve the problem of blurred work/non-work boundaries, a construct of self-control and affective factors needs to be considered because business style is changed from command to autonomy in the Smart Work context. Moreover, employees can convey their emotions easily over smartphones. Recent marketing studies have analyzed consumers' behavior based on the combination of cognitive, affective, and behavioral components, and researchers of information systems are also interested in these factors. However, previous research has some limitations, such as not classifying factors into cognitive, affective, and behavioral as well as not covering all three factors. Therefore, we explore the roles of cognitive, affective, and behavioral components in emotional exhaustion and education performance, and conduct a survey on undergraduate and graduate students, who are the major users of smartphones. Findings show that when individuals improve their cognitive capability (self-control) and usage experience (smartphone communication and internet usage), they can decrease emotional exhaustion and increase education performance. In the role of affective capability, increasing education performance is partially accepted. These results imply that organizations should not focus on controlling the usage of smartphones but on promoting appropriate smartphone usage.

• Journal of Music and Human Behavior
• /
• v.3 no.1
• /
• pp.13-28
• /
• 2006

Autistic children have difficulties in communication. They tend to have more difficulties in general expression than in understanding sentences. This thesis tested the effect of music activities on children who have difficulties in language expression. As a research method, a group of autistic middle school students was selected who were attending normal schools. Three male students were selected for the test of K-ABC and CARS. Music treatments on the selected students were applied fifteen times, twice a week, for thirty minutes every time. The total treatment process could be divided into two stages. The stage one included the first twelve treatments and the stage two included following three treatments. In the stage one, the selected autistic students learned twenty four musics under real-world-like social circumstances. In the stage two, the students repeatedly learned the twenty four musics. A week before beginning the music treatment, the students were tested by PRES and a test developed by the author of this thesis. And a week after all treatments were completed, the students were tested again with the same test method to check if the students' expressive language ability got improved. The results of the research were as the following: First, the music treatment helped improve the autistic student's ability of expressing themselves such as 'requesting', 'rejecting', 'applying social customs' and 'providing information'. Second, the author of this thesis first had a hypothesis that the tested students might show different levels of achievements according to their intellectual ability or expressive language ability in K-ABC test, PRES and CARS. But it was not true. The student's level of achievement by music activities was proved not to have a significant correlation with their intellectual or expressive language abilities in the tests. Third, it was found that, through pre- and post-test of PRES, the music treatments could improve the receptive language ability as well as the expressive language ability. It saw great effect that musical data that utilize in specially this research investigator according to children's ability, autistic children write lyrics setting in social circumstance that is revealed much routinely and composes. To front, expect that suitable a lot of musics are developed in function and ability of subject person for elevation of expressive language ability.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2005.11a
• /
• pp.268-273
• /
• 2005

기존의 텔레메틱스 서비스는 교통정보안내 및 생활정보 등의 차량 내에서의 서비스 위주로 제공한다. 그러나 차량 밖에 사용자가 있을 때 사용자가 차량의 상태 및 기타 정보를 확인하고 제어 하는 서비스의 초기단계로 전화연결 서비스(원격진단, 원격 문열림, 도난차량 추적, 내차위치확인)가 있다. 이 서비스는 중앙센터의 안내원 연결을 통하여 사용자에게 차량의 상태 및 제어를 할 수 있는 시스템이다. 본 연구는 안내원 연결을 통하지 않고 VXML을 이용하여 인간과 자동차의 인터페이스를 제공한다. VXML기술은 전화 사용자의 필요한 정보의 요구를 음성인식하며 차량으로부터의 정보는 음성합성에 의해 말로써 사용자에게 전달된다. 구현된 기술은 주차한 차량의 상태를 사용자가 궁금할 때 언제나 직접 가보지 않고 전화로 확인가능하게 함으로써 사용자의 시간적 손실을 줄인다. 또한, 주차한 차량의 훼손이나 침입 여부를 즉시 SMS로 사용자에게 알려 대처 하도록 하였다. 10인의 피실험자가 각각 10번씩 모두 100번을 실험한 결과 응답시간이 3초 이내로 나타났고 100%의 동작 성공률을 보였다. 자동차 이용자의 차량 정보 및 서비스 이용에 편리성을 제공하고, 정보 단말기와 전화 사용자 간에 VXML을 이용한 인터페이스 기술 확보 및 무한한 잠재력을 가지고 있는 지능형 자동차의 시장에 기여할 것으로 예상된다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.2
• /
• pp.362-368
• /
• 2018

Making proper judgements in urgent situations facing a submarine at the sea is very critical. This is because the commander's misjudgments could drive the entire ally to destruction in a moment. In order to generate appropriate tactics on behalf of the human commander and to analyze the effectiveness in such emergency situations, studies using intelligent agents and genetic algorithms have been conducted. In this study, inference engine based intelligent agent is adopted to each warship and submarine to generate optimal tactics on the variable environment with genetic algorithms. And we analyze the risk of the alliance according to the performance of the enemy submarine through a simple simulation and generate appropriate tactics using the genetic algorithm. Also generated tactics are evaluated and the results are analyzed to figure out why such results are formed.

• The Journal of the Korea Contents Association
• /
• v.7 no.9
• /
• pp.1-9
• /
• 2007

Recently, the primary school training courses requires creative human being who is able to solve problem in accordance with rapidly changing society. Accordingly, it needs development of edutainment contents that can develop creativity and heighten educational effect as attracting learner's interest. This paper intends to design educational game which can develop creativity. Method of research is based on the concept of creativity and theory of multiple intelligence. First, I pulled out educational elements of edutainment game which can develop ability to solve synthetic problem and then drew interest elements of edutainment game by combined game with form of cartoon. Secondly, creativity studying area set the 5 learning area of verbal, visual, mathematical, logical and analytic creativity and then, a course of learning was designed to have each 3 details of 5 teaming areas of creativity. Finally, it presented production direction of educational game by combined with 4 elements of the interest that is an avatar, achievement of a mission, a time limit and win a point.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.1
• /
• pp.144-150
• /
• 2017

Recently, deep learning is used for intelligent processing and accuracy improvement of data. It is formed calculation model composed of multi data processing layer that train the data representation through an abstraction of the various levels. A category of deep learning, convolution neural network is utilized in various research fields, which are human pose estimation, face recognition, image classification, speech recognition. When using the deep layer and lots of class, CNN that show a good performance on image classification obtain higher classification rate but occur the overfitting problem, when using a few data. So, we design the training network based on convolution neural network and trained our image data set for object classification in few class problem. The experiment show the higher classification rate of 7.06% in average than the previous networks designed to classify the object in 1000 class problem.

• The KIPS Transactions:PartB
• /
• v.8B no.5
• /
• pp.485-492
• /
• 2001

Recently, character agents are used as a user-friendly interface. In this paper, we have studied a generic framework for emotional character agents which are designed to infer emotions from diverse personalities, situations, user behaviors and to express them. The method of emotion inference is based on blackboard systems which are used to solve the problems in AI. Because it keeps independence between knowledge sources which are rules of emotions, a blackboard-based inference engine is easy to manage knowledge sources, Blackboard-based systems gave the system flexibility. So we can adapt the engine to various application systems. Each emotional agent monitors user behavior, learns user profile and infers user behavior. And it generates characters emotions according to the user profile. So, in case of same situations, the agent can generate different emotions according to users. We have studied to build an personalized emotional character agent which according to situations and user modeling.