• Proceedings of the KSME Conference
• /
• 2003.04a
• /
• pp.816-821
• /
• 2003

Researchers have utilized artificial evolution techniques and learning techniques for studying the interactions between learning and evolution. Adaptation in dynamic environments gains a significant advantage by combining evolution and learning. We propose an on-line, realtime evolutionary learning mechanism to determine the structure and the synaptic weights of a neural network controller for mobile robot navigations. We support our method, based on (1+1) evolutionary strategy which produces changes during the lifetime of an individual to increase the adaptability of the individual itself, with a set of experiments on evolutionary neural controller for physical robots behaviors. We investigate the effects of learning in evolutionary process by comparing the performance of the proposed realtime evolutionary learning method with that of evolutionary method only. Also, we investigate an interactive evolutionary algorithm to overcome the difficulties in evaluating complicated tasks.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.531-531
• /
• 2000

The challenge is to control unstable nonlinear dynamic systems using only sparse feedback from the environment concerning its performance. The design of such controllers can be achieved by evolving neural networks. An evolutionary approach to train neural networks in realtime is proposed. Evolutionary strategies adapt the weights of neural networks and the threshold values of neuron's synapses. The proposed method has been successfully implemented for pole balancing problem.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2003.06a
• /
• pp.1105-1108
• /
• 2003

Adaptation in dynamic environments gains a significant advantage by combining evolution and learning. We propose an on-line, realtime evolutionary learning mechanism to determine the structure and the synaptic weights of a neural network controller for mobile robot navigations. We support our method, based on (1+1) evolutionary strategy, which produces changes during the lifetime of an individual to increase the adaptability of the individual itself, with a set of experiments on evolutionary neural controller for physical robots behaviors.

• KIPS Transactions on Computer and Communication Systems
• /
• v.2 no.6
• /
• pp.245-252
• /
• 2013

We developed a system which supports the functions of real-time evaluation and feedback for smart learning. The system is consisted of an application for tablet PC and smart phone and the server, and the client exchanges data with the server through wireless communication. Whereby, the proposed system enabled realtime interaction and feedback between learner and teacher or between learners. As a result, the instruction for each learner's level is available using the system, and then it could enhance the level of academic achievement and learning interest.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.49 no.10
• /
• pp.883-892
• /
• 2021

This paper provides a process for developing deep learning-based aerial object detection models that can run in realtime on onboard. To improve object detection performance, we pre-process and augment the training data in the training stage. In addition, we perform transfer learning and apply a weighted cross-entropy method to reduce the variations of detection performance for each class. To improve the inference speed, we have generated inference acceleration engines with quantization. Then, we analyze the real-time performance and detection performance on custom aerial image dataset to verify generalization.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.9
• /
• pp.267-280
• /
• 2020

Mirror is a physical reflective surface, typically of glass coated with a metal amalgam, and it is to reflect an image clearly. They are available everywhere anytime and become an essential tool for us to observe our faces and appearances. With the advent of modern software technology, we are motivated to enhance the reflection capability of mirrors with the convenience and intelligence of realtime processing, microservices, and machine learning. In this paper, we present a development of Digital Mirror System that provides the realtime reflection functionality as mirror while providing additional convenience and intelligence including personal information retrieval, public information retrieval, appearance age detection, and emotion detection. Moreover, it provides a multi-model user interface of touch-based, voice-based, and gesture-based. We present our design and discuss how it can be implemented with current technology to deliver the realtime mirror reflection while providing useful information and machine learning intelligence.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.48 no.1
• /
• pp.51-59
• /
• 2011

In this paper, we propose the standardized idea for learning environment of e-Learning 2.0 (originated from Web 2.0 paradigm) to overcome limitations of current passive e-Learning environment. Because current LMS/ LCMS which manages e-learning systems has the limitation of providing various interactive elements when people use video contents to study, it lacks real-timeness and interactivity between teacher-learner and learner-learner in the operation of video contents and has difficulty in measuring accurate progress rate of learners in the process of teaching-learning. Therefore, we designed multimedia contents(both-way learning requisite) to overcome limitations of current e-Learning system and to maximize the effect of learning of learners so that it makes possible to interact between learners and teachers in realtime. For this, this thesis designs the standardized idea based on expanded SCORM standard which is now used for manufacturing LMS/LCMS, and according to ideas we have proposed, it implements e-learning multimedia environment.

• Journal of Korea Multimedia Society
• /
• v.21 no.5
• /
• pp.599-608
• /
• 2018

As natural disasters are increasing due to the unusual weather and the modern society is getting complicated, the rapid change of the urban environment has increased human disasters. Thus, citizens are becoming more anxious about social safety. The importance of preparation for safety has been suggested by providing the disaster safety services such as regional safety index, life safety map, and disaster safety portal application. In this paper, we propose an application framework to predict the urban safety index based on user's location with realtime weather/atmosphere data after creating a predication model based on the machine learning using number of occurrence cases and weather/atmosphere history data. Also, we implement an application to provide traffic safety index with executing preprocessing occurrence cases of traffic and weather/atmosphere data. The existing regional safety index, which is displayed on the Si-gun-gu area, has been mainly utilized to establish safety plans for districts vulnerable to national policies on safety. The proposed system has an advantage to service useful information to citizens by providing urban safety index based on location of interests and current position with realtime related data.

• The Journal of Information Systems
• /
• v.6 no.2
• /
• pp.165-188
• /
• 1997

This paper describes about the development of the 3 Dimensional interactive physics experiment system using virtual reality technologies. Until now, most of the computer aided education systems have adopted one-sided learning way which only shows or tells. It excludes students during learning processes. To solve the problems of the one-sided learning way and to improve the educational productivity, the 3 Dimensional interactive physics experiment system is developed. The 3 Dimensional interactive physics experiment system introduced in this paper provides a new learning motivation for students and improves their educational effects through the 3 Dimensional graphics, realtime action, and realistic interactive experiment.

• Journal of the Korean Society of Radiology
• /
• v.10 no.8
• /
• pp.637-643
• /
• 2016

Due to the popularity of artificial intelligent, medical image processing using artificial neural network is increasingly attracting the attention of academic and industry researches. Deep learning with a convolutional neural network has been proved to very effective representation of images. However, the training process requires high performance H/W platform. Thus, the realtime learning of a large number of high dimensional samples within low-power devices is a challenging problem. In this paper, we attempt to establish this possibility by presenting a realtime neural network method on Raspberry pi using online sequential extreme learning machine. Our experiments on high-dimensional dataset show that the proposed method records an almost real-time execution.