• Review of KIISC
• /
• v.33 no.4
• /
• pp.31-40
• /
• 2023

차량 기술의 발전으로 커넥티드 및 자율주행 차량 환경과 같은 차량 기술의 발전은 운전자에게 편의와 안전을 위한 기능을 제공한다. 하지만 이러한 장점들에도 불구하고 사이버 공격의 다양한 취약점 노출되어 있다. 최근까지 자동차 내부 네트워크로써 가장 널리 사용되는 통신기술인 CAN(Controller Area Networks) 통신은 대부분 차량의 동력을 담당하는 역할을 하다보니, 보안 문제의 중심에 서게 될 수 있다. 다양한 이기종 네트워크에서 구성된 가상화 기반의 사이버 훈련 프레임워크에 대해 설명하고자 한다. 이러한 사례 연구는 차량에 대해 물리적인 실험 환경에서의 모의침투와 같은 테스트가 어렵고, 보안과 안전이 함께 고려되어야 하는 특수성을 가진 차량 내부 네트워크의 사이버 훈련 프레임워크 설계에 도움을 줄 것이다. 본 논문에서는 가상화 개발환경 구축 사례 조사, 가상화 개발환경 구현과 차량의 공격 시나리오 및 탐지 프레임워크의 동향을 설명한다.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.14 no.5
• /
• pp.969-976
• /
• 2019

A lightweight artificial intelligence hardware has made great strides in many application areas. In general, a lightweight artificial intelligence system consist of lightweight artificial intelligence engine and preprocessor including feature selection, generation, extraction, and normalization. In order to achieve optimal performance in broad range of applications, lightweight artificial intelligence system needs to choose a good preprocessing function and set their respective hyper-parameters. This paper proposes a unified framework for a lightweight artificial intelligence system and utilization method for finding models with optimal performance to use on a given dataset. The proposed unified framework can easily generate a model combined with preprocessing functions and lightweight artificial intelligence engine. In performance evaluation using handwritten image dataset and fall detection dataset measured with inertial sensor, the proposed unified framework showed building optimal artificial intelligence models with over 90% test accuracy.

• Journal of The Korean Association of Information Education
• /
• v.24 no.5
• /
• pp.453-462
• /
• 2020

With the development of artificial intelligence, a lot of platforms have emerged that enable anyone to easily access and learn about artificial intelligence or create artificial intelligence models. Therefore, in this study, we analyzed various artificial intelligence education platforms and developed and proposed a SW education class plan using a framework-based artificial intelligence education platform for activating artificial intelligence based SW education. The artificial intelligence-based SW education framework aims to cultivate artificial intelligence literacy on the basis of computational thinking. In addition, a learner-centered project class was formed to include elements that could be fused with real life contexts or other subjects. Using this, with the theme of creating an artificial intelligence program to help separate garbage collection, a six-hour project-based class was developed and proposed using practical arts, social studies, and creative experiential activities. This project class was organized using a platform that is not difficult, such as AI Oceans and Entry.

• Journal of The Korean Association of Information Education
• /
• v.23 no.6
• /
• pp.639-653
• /
• 2019

The purpose of this study is to design an instructional framework and cognitive learning environment for AI education based on computational thinking in order to ground the theoretical rationale for AI education. Based on the literature review, the learning model is proposed to select the algorithms and problem-solving models through the abstraction process at the stage of data collection and discovery. Meanwhile, the instructional model of AI education through computational thinking is suggested to enhance the problem-solving ability using the AI by performing the processes of problem-solving and prediction based on the stages of automating and evaluating the selected algorithms. By analyzing the research related to the cognitive learning environment for AI education, the instructional framework was composed mainly of abstraction which is the core thinking process of computational thinking through the transition from the stage of the agency to modeling. The instructional framework of AI education and the process of constructing the cognitive learning environment presented in this study are characterized in that they are based on computational thinking, and those are expected to be the basis of further research for the instructional design of AI education.

• Review of KIISC
• /
• v.33 no.4
• /
• pp.23-29
• /
• 2023

대부분의 임베디드 시스템은 기계장치와 전자기기 장치가 함께 작동되는 물리 장치로써, 이기종 네트워크, 복잡한 보안체계 등을 고려하여 가상화 기반 사이버훈련 환경이 구성되어야 한다. 또한, 차량을 대상으로 물리적인 실험환경에서 모의침투 등 사이버훈련을 수행한다는 것은 교통사고를 비롯한 안전사고 발생에 있어 위험이 존재한다. 본 논문에서는 가상 개발환경에서의 공격 기반 차량용 사이버훈련 프레임워크를 제안하고자 한다. 먼저, 공격 기반 차량용 사이버훈련 프레임워크의 작동은 자동 활성화되는 가상의 CAN 네트워크 인터페이스로 시작된다. 가상의 CAN 네트워크 인터페이스는 가상 머신에서 간단한 부트스트랩 명령어 실행을 통해 파이썬 패키지와 Ubuntu 서비스 목록 설치 명령이 자동으로 실행되면서 설치된다. 이후 내부 네트워크 시뮬레이터와 공격모듈과 관련된 UI가 자동으로 Ubuntu Systemd에 의해 백그라운드에서 실행되어 시작과 동시에 준비 상태를 유지하게 된다. 사이버훈련 UI 내 공격 모듈은 사용자에 의한 공격 선택 및 파라미터 셋팅 이후 차량의 이상 상태를 사이버훈련 UI에 다시 출력되게 된다. 본 논문에서 제안하는 가상 개발환경 기반의 차량용 사이버훈련 프레임워크는 자율주행 차량 사고의 위험이나 다른 특수한 제약 없이 사용자의 학습 경험을 확장시킬 수 있다. 또한, 기존의 가상화 기반 사이버훈련 교육 콘텐츠와는 달리 일반 사용자들이 접근하기 쉬운 형태로 확장 개발이 가능하다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.11a
• /
• pp.310-313
• /
• 2017

범죄수사 과정에서 많은 양의 데이터를 시간 내에 분석하는 것은 성공적인 포렌식의 필수 요소이다. 컴퓨터와 사람 모두에게 있어, 시간과 자원의 제한은 수사 결과에 부정적인 영향을 가져온다. 그러므로 현재 사용되고 있는 다양한 포렌식 도구에는 시간과 자원의 효율적인 사용이 필요하다. 사례기반추론 및 멀티에이전트 시스템과 같은 인공지능 기반의 도구를 통해 디지털 포렌식 수사를 효과적으로 도울 수 있다. 본 논문에서는 인공지능을 활용한 지능형 포렌식 도구 및 프레임워크를 분석하고, 오늘날의 프레임워크의 한계점과 미래에 관해 논의한다. 인공지능 기반 시스템의 목적은 수사에서의 증거를 포함한 데이터를 분석하고 연관성을 밝힘으로서 포렌식 전문가에게 중요한 단서를 제공하고 직접 분석해야하는 데이터의 양을 줄이는 것에 있다. 이러한 인공지능의 활용은 많은 양의 데이터를 수사할 때 사람이 간과할 수 있는 증거들을 연결시켜주는 데에 큰 도움이 된다.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.12 no.1
• /
• pp.115-122
• /
• 2017

Deep learning is artificial intelligence technology that can teach people like themselves who need machine learning. Deep learning has become of the most promising in the development of artificial intelligence to understand the world and detection technology, and Google, Baidu and Facebook is the most developed in advance. In this paper, we discuss the kind of deep learning frameworks, compare and analyze the efficiency of the image and speech recognition field of it.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.10 no.9
• /
• pp.2361-2368
• /
• 2009

Recently AI(Artificial Intelligence) is one of the issues in the on-line game, a research that a game character seems to be realistic and is progressing using AI technique. Especially NPC is an important part of the AI researches of on-line game, and it is concerned by a game player and an architect. We proposed an intelligent agent framework to implement the NPC technique after studying the NPC technique using context awareness that reacts to the PC(Player Character) actively. Also, it can be developed gradually, and apply to various application because it has the capability to of adding an agent or deleting an agent easily.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.274-276
• /
• 2021

As artificial intelligence develops rapidly, we can experience it in our everyday life such as with medical, education, and game applications. Traditional SW services were programmed explicitly by the intention of the programmer, and we have conducted evaluation on it. However, due to the uncertianty of AI services, risk follows to the products. Therefore, UX evaluations need to be different from traditional UX evaluations. Therefore, in this paper we suggest a AI-UX framework that consideres the task delegability, UX evaluations metrics, and individual differences.

• Journal of Software Assessment and Valuation
• /
• v.17 no.1
• /
• pp.101-113
• /
• 2021

In this paper, we introduce a speaker overlap system and look at the process of converting the existed system on the specific framework of artificial intelligence. Speaker overlap is when two or more speakers speak at the same time during a conversation, and can lead to performance degradation in the fields of speech recognition or speaker recognition, and a lot of research is being conducted because it can prevent performance degradation. Recently, as application of artificial intelligence is increasing, there is a demand for switching between artificial intelligence frameworks. However, when switching frameworks, performance degradation is observed due to the unique characteristics of each framework, making it difficult to switch frameworks. In this paper, the process of converting the speaker overlap detection system based on the Keras framework to the pytorch-based system is explained and considers components. As a result of the framework switching, the pytorch-based system showed better performance than the existing Keras-based speaker overlap detection system, so it can be said that it is valuable as a fundamental study on systematic framework conversion.