• Proceedings of the KSPS conference
• /
• 2000.07a
• /
• pp.17-28
• /
• 2000

Recently speech technologies have shown notable advancements and they now play major roles in computer-aided language learning systems. In the current paper, use of speech recognition technologies is viewed with our system for teaching English pronunciation to Japanese speakers.

• Journal of The Korean Association For Science Education
• /
• v.22 no.2
• /
• pp.299-313
• /
• 2002

A web-based instruction program for the enriched course under the 7th Revised National Curriculum of Biology in Korea was developed and the application effects to learners were analyzed. For the development of the web-based instruction program, five topics of biology from the enriched courses through 7th to 10th grades in the middle and high school science textbooks were selected and modulated with interrogative sentences. Each topic of programs was divided into four activity sections according to the learners' activity procedures supplemented with explanations and evaluations. Each activity was hyper-linked to multi-layers and animations. Further, a virtual experiment was also developed and an evaluation section designed by Java Script was attached. Among five topics, one topic of 'Reproduction and development' at 9th grade level was selected to examine the effects on students' learning. Among 247 9th grade students in the research subject school, only 67 students were able to accessible to ultra-thin Internet cables with their computers at home and they became an experimental group. A control group was assigned to those who are similar level of school science achievement to the experiment group and did not use the web-based program. It was found that most of 9th grade students are able to use Internet at home, however, they do not prefer to use Internet for homework or task project. Rather, most of students used Internet for e-mail or information navigation. Students used internet to solve problems of science and perceived the benefits of Internet for science learning. However, there are not many students to utilize Internet for science homework or task project. Students expressed that they do not prefer to use a web-based learning program for science learning due to lack of interests in science. The effects on students who studied with this program appeared to be significantly high compared to those who did not study with this program. Students who studied with this program positively evaluated this program, in particular, they enjoyed animation effect and virtual experiments. It was concluded that a web-based program for science learning should be developed and distributed through Internet in an attractive and interesting format for students. It was also concluded that various web-based programs for science learning with animation effect and virtual experiments should be developed to increase students' interests in science as well as to improve students' science achievements.

• Journal of the Korea Society of Computer and Information
• /
• v.3 no.2
• /
• pp.125-130
• /
• 1998

One of the most urgent problems facing 21 century might seem to be a preparation for an advanced education against an information society and have relation to a new information supporting system in teaching as well as leaning. Nevertheless, our educational circumstances have a lot of problems because an effective information supporting contents for teaching and leaning is insufficient. Especially, we've never been making every effort to develope educational contents providing students with a type of information and knowledge database even though the education which come into use internet has been raising a social issue. Therefore, we propose a model of learning contents for supporting www based distance learning system.

• ETRI Journal
• /
• v.28 no.1
• /
• pp.59-66
• /
• 2006

Learning behaviors of a radial basis function network (RBFN) using a singular value decomposition (SVD) and stochastic gradient (SG) algorithm, together named RBF-SVD-SG, for odor sensing systems are analyzed, and a fast training method is proposed. RBF input data is from a conducting polymer sensor array. It is revealed in this paper that the SG algorithm for the fine-tuning of centers and widths still shows ill-behaving learning results when a sufficiently small convergence coefficient is not used. Since the tuning of centers in RBFN plays a dominant role in the performance of RBFN odor sensing systems, our analysis is focused on the center-gradient variance of the RBFN-SVD-SG algorithm. We found analytically that the steadystate weight fluctuation and large values of a convergence coefficient can lead to an increase in variance of the center-gradient estimate. Based on this analysis, we propose to use the least mean square algorithm instead of SVD in adjusting the weight for stable steady-state weight behavior. Experimental results of the proposed algorithm have shown faster learning speed and better classification performance.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.15 no.2
• /
• pp.51-60
• /
• 2020

Optical character recognition (OCR) has been studied for decades because it is very useful in a variety of places. Nowadays, OCR's performance has improved significantly due to outstanding deep learning technology. Thus, there is an increasing demand for commercial-grade but affordable OCR systems. We have developed a low-cost, high-performance OCR system for the industry with the cheapest embedded developer kit that supports GPU acceleration. To achieve high accuracy for industrial use on limited computing resources, we chose a state-of-the-art text recognition algorithm that uses an end-to-end deep learning network as a baseline model. The model was then improved by replacing the feature extraction network with the best one suited to our conditions. Among the various candidate networks, EfficientNet-B3 has shown the best performance: excellent recognition accuracy with relatively low memory consumption. Besides, we have optimized the model written in TensorFlow's Python API using TensorFlow-TensorRT integration and TensorFlow's C++ API, respectively.

• The Journal of Korea Robotics Society
• /
• v.16 no.3
• /
• pp.260-269
• /
• 2021

To verify performance or conduct experiments using actual robots, a lot of costs are needed such as robot hardware, experimental space, and time. Therefore, a simulation environment is an essential tool in robotics research. In this paper, we develop the HUMIC simulator using ROS and Gazebo. HUMIC is a humanoid robot, which is developed by HCIR Lab., for human-robot interaction and an upper body of HUMIC is similar to humans with a head, body, waist, arms, and hands. The Gazebo is an open-source three-dimensional robot simulator that provides the ability to simulate robots accurately and efficiently along with simulated indoor and outdoor environments. We develop a GUI for users to easily simulate and manipulate the HUMIC simulator. Moreover, we open the developed HUMIC simulator and GUI for other robotics researchers to use. We test the developed HUMIC simulator for object detection and reinforcement learning-based navigation tasks successfully. As a further study, we plan to develop robot behavior intelligence based on reinforcement learning algorithms using the developed simulator, and then apply it to the real robot.

• Nuclear Engineering and Technology
• /
• v.56 no.4
• /
• pp.1472-1479
• /
• 2024

Cyber-Physical Energy Systems (CPESs) integrate cyber and hardware components to ensure a reliable and safe physical power production and supply. Renewable Energy Sources (RESs) add uncertainty to energy demand that can be dealt with flexible operation (e.g., load-following) of CPES; at the same time, scenarios that could result in severe consequences due to both component stochastic failures and aging of the cyber system of CPES (commonly overlooked) must be accounted for Operation & Maintenance (O&M) planning. In this paper, we make use of Deep Reinforcement Learning (DRL) to search for the optimal O&M strategy that, not only considers the actual system hardware components health conditions and their Remaining Useful Life (RUL), but also the possible accident scenarios caused by the failures and the aging of the hardware and the cyber components, respectively. The novelty of the work lies in embedding the cyber aging model into the CPES model of production planning and failure process; this model is used to help the RL agent, trained with Proximal Policy Optimization (PPO) and Imitation Learning (IL), finding the proper rejuvenation timing for the cyber system accounting for the uncertainty of the cyber system aging process. An application is provided, with regards to the Advanced Lead-cooled Fast Reactor European Demonstrator (ALFRED).

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.2
• /
• pp.147-153
• /
• 2024

X-rays are extensively employed for non-destructive inspection, applied to packaged food, human anatomy, and industrial products. Recently, this technology has extended to inspecting batteries in electric vehicles. Given the challenge of manual inspection for a substantial volume of batteries, deep learning is leveraged to detect battery defects. However, the effectiveness of deep learning heavily depends upon data size, and acquiring authentic defective images is a difficult and time-consuming task. In this study, we use data augmentation and investigate the impact of data size on battery inspection performance. The results provide valuable insights for enhancing the capabilities of the inspection process.

• International conference on construction engineering and project management
• /
• 2024.07a
• /
• pp.769-774
• /
• 2024

The availability of PMT(Project Management Tool) in the market has been increasing rapidly in recent years and Significant advancements have been made for project managers to use for planning, monitoring, and control. Recently, studies applying the Reinforcement-Learning Based Construction Schedule Simulation algorithm for construction project process planning/management are increasing. When reinforcement learning is applied, the agent recognizes the current state and learns to select the action that maximizes the reward among selectable actions. However, if the action of global optimal points is not selected in simulation selection, the local optimal resource may receive continuous compensation (+), which may result in failure to reach the global optimal point. In addition, there is a limitation that the optimization time can be long as numerous iterations are required to reach the global optimal point. Therefore, this study presented a method to improve optimization performance by increasing the probability that a resource with high productivity and low unit cost is selected, preventing local optimization, and reducing the number of iterations required to reach the global optimal point. In the performance evaluation process, we demonstrated that this method leads to closer approximation to the optimal value with fewer iterations.

• Journal of Sensor Science and Technology
• /
• v.33 no.5
• /
• pp.298-304
• /
• 2024

In Korea, fires are the second most common type of disaster, causing large-scale damages. The installation of fire detectors is legislated to prevent fires and minimize damage. Conventional fire detectors have limitations in initial suppression of failures because they detect fires when large amounts of smoke and heat are generated. Additionally, frequent malfunctions in fire detectors may cause users to turn them off. To address these issues, recent studies focus on accurately detecting even small-scale fires using multi-sensor and deep-learning technologies. They also aim at quick fire detection and thermal decomposition using gas. However, these studies are not practical because they overlook the heavy computations involved. Therefore, we propose a fast and accurate fire detection system based on multi-sensor and deep-learning technologies. In addition, we propose a computation-reduction method for selecting sensors suitable for detection using the Pearson correlation coefficient. Specifically, we use a moving average to handle outliers and two-stage labeling to reduce false detections during preprocessing. Subsequently, a deep-learning model is selected as LSTM for analyzing the temporal sequence. Then, we analyze the data using a correlation analysis. Consequently, the model using a small data group with low correlation achieves an accuracy of 99.88% and a false detection rate of 0.12%.