• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.17 no.2
• /
• pp.113-120
• /
• 2024

The population sparseness index(PSI) is being utilized to describe the functioning of internal layers in artificial neural networks from the perspective of neurons, shedding light on the black-box nature of the network's internal operations. There is research indicating a positive correlation between the PSI and performance in each layer of convolutional neural network models for image classification. In this study, we observed the internal operations of a convolutional neural network when adversarial examples were applied. The results of the experiments revealed a similar pattern of positive correlation for adversarial examples, which were modified to maintain 5% accuracy compared to applying benign data. Thus, while there may be differences in each adversarial attack, the observed PSI for adversarial examples demonstrated consistent positive correlations with benign data across layers.

• Journal of Digital Contents Society
• /
• v.18 no.8
• /
• pp.1627-1633
• /
• 2017

Human dance is difficult to learn since there is no effective way to imitate an expert's motion, a sequence of complicated body movements, without taking an actual class. In this paper, we propose a dance learning system using human depth information. In the proposed system, a set of example motions are captured from various expert dancers through a marker-free motion capture and archived into a motion database server for online dance lessons. Given the end-user devices such as tablet and kiosk PCs, a student can learn a desired motion selected from the database and send one's own motion to an instructor for online feedback. During this learning process, our system provides a posture-based motion search and multi-mode views to support the efficient exchange of motion data between the student and instructor under a networked environment. The experimental results demonstrate that our system is capable to improve the student's dance skills over a given period of time.

• Journal of the Korea Computer Graphics Society
• /
• v.27 no.1
• /
• pp.9-17
• /
• 2021

In this paper, we present a simple and fast supervised learning framework based on model predictive control so as to learn motion controllers for a physic-based character to track given example motions. The proposed framework is composed of two components: training data generation and offline learning. Given an example motion, the former component stochastically controls the character motion with an optimal controller while repeatedly updating the controller for tracking the example motion through model predictive control over a time window from the current state of the character to a near future state. The repeated update of the optimal controller and the stochastic control make it possible to effectively explore various states that the character may have while mimicking the example motion and collect useful training data for supervised learning. Once all the training data is generated, the latter component normalizes the data to remove the disparity for magnitude and units inherent in the data and trains an artificial neural network with a simple architecture for a controller. The experimental results for walking and running motions demonstrate how effectively and fast the proposed framework produces physics-based motion controllers.

• Journal of Practical Engineering Education
• /
• v.8 no.2
• /
• pp.95-101
• /
• 2016

In this paper, we introduce an implementation method of switch-based LED (Light Emitting Diode) Art logic circuits to help understanding the operation principle of digital logic circuits. Digital logic circuit practice using bread board is widely practiced in colleges or high schools in South Korea. However, actual digital logic circuit practice lacks examples of basic implementation, and as results of this problem, study with more complicated examples disturbs understanding the basic operation principle of digital logic circuits. Therefore, we proposed and tested an implementation method of switch-based LED Art logic circuits to help understanding the necessity of digital logic circuits which control signals of multiple output devices simultaneously.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.16 no.1
• /
• pp.1-7
• /
• 2023

Convolutional neural networks have already been applied to various fields beyond human visual processing capabilities in the image processing area. However, they are exposed to a severe risk of deteriorating model performance due to the appearance of adversarial attacks. In addition, defense technology to respond to adversarial attacks is effective against the attack but is vulnerable to other types of attacks. Therefore, to respond to an adversarial attack, it is necessary to analyze how the performance of the adversarial attack deteriorates through the process inside the convolutional neural network. In this study, the adversarial attack of the Alexnet and VGG11 models was analyzed using the population sparseness index, a measure of neuronal activity in neurophysiology. Through the research, it was observed in each layer that the population sparsity index for adversarial examples showed differences from that of benign examples.

• Journal of the Korea Computer Graphics Society
• /
• v.28 no.2
• /
• pp.29-41
• /
• 2022

In this paper, we introduce a novel real-time, interactive animation system which uses real-time motion inputs from a low-cost motion-sensing device Kinect. Our system generates interaction motions between the user character and the counterpart character in real-time. While the motion of the user character is generated mimicking the user's input motion, the other character's motion is decided to react to the user avatar's motion. During a pre-processing step, our system analyzes the reference motion data and generates mapping model in advance. At run-time, our system first generates initial poses of two characters and then modifies them so that it could provide plausible interacting behavior. Our experimental results show plausible interacting animations in that the user character performs a modified motion of user input and the counterpart character properly reacts against the user character. The proposed method will be useful for developing real-time interactive animation systems which provide a better immersive experience for users.

• Journal of the Korea Computer Graphics Society
• /
• v.11 no.2
• /
• pp.9-15
• /
• 2005

가상환경에서의 실시간 캐릭터 애니메이션은 컴퓨터 게임 및 가상현실 분야에서 중요한 문제이다. 최근에는 캐릭터 애니메이션의 사실성을 높이기 위해 대규모 동작 캡쳐 데이터를 활용하고 있다. 그러나 거대하고 복잡한 가상환경을 만들기 위해서는 현실 환경에서 동작을 캡쳐하고 이를 넓은 가상환경에 적용해야하는 어려움이 따른다. 본 논문에서는 애니메이션 캐릭터가 몇 가지 빌딩 블록(building block)들로 건설된 넓은 가상공간에서 자유롭게 활동하게 하는 새로운 방법을 제안한다. 빌딩 블록들은 반복적인 형태로 배열되어 더 큰 환경을 생성할 수 있다. 우리는 이러한 각각의 블록들을 모션 패치(motion patch)라고 부른다. 모션 패치는 해당블록 내부에서 캐릭터가 수행할 수 있는 동작들에 대한 정보를 가지고 있다. 우리는 다수의 캐릭터가 방대하고 복잡한 가상환경 속에서 실시간으로 애니메이션되고 제어되는 예제들을 통해 이 논문에서 제시하는 방법이 다양한 목적에 유용하게 사용될 수 있음을 보인다.

• Proceedings of the Korea Society for Simulation Conference
• /
• 2002.11a
• /
• pp.79-83
• /
• 2002

High Level Architecture (HLA)는 이기종 시뮬레이터 간의 연동을 위한 미들웨어이다. 본 논문에서는 Discrete Event System Specification (DEVS) 형식론을 바탕으로 한 DEVS 모델의 시뮬레이션을 HLA 기반에서 수행하는 방법에 대하여 제시한다. DEVS 시뮬레이션 메시지를 HLA의 서비스를 사용하여 변환하여 기존의 시뮬레이션 알고리즘을 그대로 사용하여 분산 시뮬레이션 할 수 있는 방식에 대하여 기술한다. 또한 모델 사이의 연결 정보를 HLA 환경에서 이용하는 방식에 대하여 기술한다. 제안된 방식을 구현한 DEVSimHLA 시뮬레이션 환경의 구조 및 동작 방식에 대하여 알아보고 이를 이용한 간단한 예제를 보인다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06b
• /
• pp.494-499
• /
• 2008

임베디드 시스템이 주목받으면서 개발상의 편의를 위해 시스템 시뮬레이터가 다양한 용도로 사용되고 있다. 시스템이 복잡해지고 소프트웨어의 규모가 커지면서 이러한 시스템 시뮬레이터들에 있어 그 성능은 매우 중요한 이슈가 되고 있는데, 본 논문에서는 공유 메모리를 사용하여 통신하는 다중 프로세서 시스템에서 동기화 횟수를 줄이는 방법을 제안하고 이를 기반으로 한 다중 프로세서 시스템 시뮬레이터를 개발하였다. 이 시뮬레이터는 프로세서 시뮬레이터의 내부를 크게 고치지 않고 공유 메모리 접근만을 가로채 동작이 가능하므로 쉽게 다양한 종류의 프로세서를 연결할 수 있는 확장성 역시 가지고 있다. 제안하는 동기화 기법과 개발된 시뮬레이터는 7개의 프로세서를 사용하여 동작하는 JPEG 인코더 예제의 구동을 통해 테스트되었으며, 이를 통해 인과율을 깨뜨리지 않고도 빠른 시뮬레이션이 가능함을 확인할 수 있었다.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.45 no.4
• /
• pp.1-10
• /
• 2008

The problem of controlling asynchronous sequential machines is addressed in this paper. Corrective control means to make behavior of an asynchronous sequential machine equal to that of a given model. The main objective is to develope a corrective controller, especially when a model is given as nondeterministic, or a set of reference models. The structure of corrective control system for asynchronous sequential machines is addressed first, followed by description of nondeterministic models. Then, we propose a method for analyzing reachability of asynchronous machines and nondeterministic models. Proposed methods are demonstrated in an example.