• The Journal of Korean Institute of Next Generation Computing
• /
• v.14 no.5
• /
• pp.31-39
• /
• 2018

Recently, there has been active studies to provide a user-friendly interface in a virtual reality environment by recognizing user hand gestures based on deep learning. However, most studies use separate sensors to obtain hand information or go through pre-process for efficient learning. It also fails to take into account changes in the external environment, such as changes in lighting or some of its hands being obscured. This paper proposes a hand gesture recognition method based on deep learning that is strong in external environments without the need for pre-process of RGB images obtained from general webcam. In this paper we improve the VGGNet and the GoogLeNet structures and compared the performance of each structure. The VGGNet and the GoogLeNet structures presented in this paper showed a recognition rate of 93.88% and 93.75%, respectively, based on data containing dim, partially obscured, or partially out-of-sight hand images. In terms of memory and speed, the GoogLeNet used about 3 times less memory than the VGGNet, and its processing speed was 10 times better. The results of this paper can be processed in real-time and used as a hand gesture interface in various areas such as games, education, and medical services in a virtual reality environment.

• The Journal of Korea Robotics Society
• /
• v.17 no.3
• /
• pp.264-272
• /
• 2022

Human-robot cooperative tasks are increasingly required in our daily life with the development of robotics and artificial intelligence technology. Interactive reinforcement learning strategies suggest that robots learn task by receiving feedback from an experienced human trainer during a training process. However, most of the previous studies on Interactive reinforcement learning have required an extra feedback input device such as a mouse or keyboard in addition to robot itself, and the scenario where a robot can interactively learn a task with human have been also limited to virtual environment. To solve these limitations, this paper studies training strategies of robot that learn table balancing tasks interactively using deep reinforcement learning with human's facial expression feedback. In the proposed system, the robot learns a cooperative table balancing task using Deep Q-Network (DQN), which is a deep reinforcement learning technique, with human facial emotion expression feedback. As a result of the experiment, the proposed system achieved a high optimal policy convergence rate of up to 83.3% in training and successful assumption rate of up to 91.6% in testing, showing improved performance compared to the model without human facial expression feedback.

• KIISE Transactions on Computing Practices
• /
• v.24 no.3
• /
• pp.122-128
• /
• 2018

In recent years, virtual training has attracted a lot of attention because it is as effective as traditional in-person training and provides more cost-effective and safer learning environment. However, the existing virtual training systems rely on the evaluator's qualitative judgement when evaluating and analyzing trainees' performance. Additionally, evaluation and analysis functions are only available on certain systems so those functions need to be developed for each system. In this paper, we propose an extensible evaluation and analysis system for virtual training using experiential knowledge collected from experts for providing effective evaluation and analysis. Specifically, we provide a method of applying Open API so that the proposed system works with different types of virtual training system. In addition, the experiential knowledge is constructed in advance for the evaluation and analysis so that the efficiency of the evaluation with the comparison target is increased. This experiential knowledge can be quantitatively compared to trainees' performance according to the proposed evaluation and analysis procedures.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.4
• /
• pp.541-547
• /
• 2022

Realistic education using virtual reality compared to traditional learning can enhance students' understanding of knowledge through immersion and interaction. In previous studies, VR education is mainly focused on experiences, and it is difficult to find its applications for environmental education. Environmental issues are a global problem, and environmental education is essential for the future. In this research, we developed an immersive virtual reality-based environmental education simulation designed to help students recognize the importance of environmental education and participate in environmental-friendly actions. This simulation is based on the virtual ecosystem model, which maintains a casual relationship among environmental factors, spatio-temporal connection, and persistent state. Users intuitively recognize environmental problems and is motivated to solve the problem while experiencing the results of interaction related to environmental factors in virtual environment.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.937-944
• /
• 2006

Although people's anxiety about the environmental problem has been getting higher, they are not provided good quality of knowledge about the environment. Based on this situation, Ecoblog can be a new type of online community to educate the public in ecological knowledge. Especially, Ecoblog can be utilized as a method of "preventive education", and it will contribute to reduce great amounts of environmental budget to restore contaminated environment to previous condition. Ecoblog also utilizes the concept of blog which user can create and append their site with chosen themes. A weblog or a blog is a non-commercial webpage regularly updated through the use of a blogging software which allows the user to "publish" kinds of amalgamations of text and graphics to the page as posts. The technology offered in Ecoblog is utilizing the concept of 4D place and game metaphor in order to provide users the sense of participation, interaction and immersion among them and the growing community. Thus, it requires applying the CAAD technology by implementing semantically well-defined building data model as a core database to create a 4D virtual community. This research focuses on defining a 4d spatial framework suitable for developing an online ecological community. Through our study, the state-of-the-art of online community has been studied at the first step. Second, the scenario of using EcoBlog described with content, visualization and navigation are defined based on the critical features derived at the first step. Finally, a 4d spatial framework composed of semantic building data model, content and rule database is constructed to propose factors that are necessary to establish an ecological virtual community. In conclusion, our framework could enhance the comprehension and interaction between users and virtual buildings in the ecological community by integrating the concept of game design, 4D CAD and semantic data model. Such framework can be applied to any online community for an educational purpose.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.49-54
• /
• 2008

In order to do a maintenance task, a maintenance operator should learn the basic skills of the maintenance task such as assembly and disassembly (A/D). However, the key of the learning process is to learn the A/D task intuitively and naturally. Haptic guidance promises to give effectiveness and benefit qualitatively since a person can be trained to do the optimal task based on information that comes from an expert, database, or intelligent algorithms. By applying haptic guidance, a maintenance training process can be made more intuitive and natural in a virtual environment. This paper describes the development of a maintenance training system by using haptic guidance.

• Journal of the Korea Society for Simulation
• /
• v.5 no.2
• /
• pp.13-24
• /
• 1996

Workflow is an ordered sequence of interdependent component data activities each of which can be executed on an integrated information system by accessing a remote information system. In our previous research [4], we proposed a distributed CIM Workflow system which consists of a workflow execution model called DAF-Net and an agent-based information systems called AIMIS. Given a component data activity, there needs an interaction protocol among agents which allocates the component data activity to a relevant information systems exist. The objective of this research is to propose and test two protocols: ARR(Asynchronous Request and Response)protocol and NCL(Negotiation with Case based Learning) protocol. To test the effectiveness of the protocols, we applied the PVM(Parallel Virtual Machine) software to simulate the distributed CIM Workflow system. PVM provides a distributed computing environment in which users can run different software processes in different computers while allowing communication among the processes.

• Journal of Korea Multimedia Society
• /
• v.25 no.2
• /
• pp.197-205
• /
• 2022

The Cyber Physical System(CPS) is an important concept in achieving SMSs(Smart Manufacturing Systems). Generally, CPS consists of physical and virtual elements. The former involves manufacturing devices in the field space, whereas the latter includes the technologies such as network, data collection and analysis, security, and monitoring and control technologies in the cyber space. Currently, all these elements are being integrated for achieving SMSs in which we can control and analyze various kinds of producing and diagnostic issues in the cyber space without the need for human intervention. In this study, we focus on implementing a production equipment monitoring system related to building a SMS. First, we describe the development of a fog-based gateway system that links physical manufacturing devices with virtual elements. This system also interacts with the cloud server in a multimedia network environment. Second, we explain the proposed network infrastructure to implement a monitoring system operating on a cloud server. Then, we discuss our monitoring applications, and explain the experience of how to apply the ML(Machine Learning) method for predictive diagnostics.

• Journal of Internet Computing and Services
• /
• v.22 no.5
• /
• pp.27-33
• /
• 2021

With the broad adoption of the Internet of Things (IoT) in a variety of scenarios and application services, management and orchestration entities require upgrading the traditional architecture and develop intelligent models with ultra-reliable methods. In a heterogeneous network environment, mission-critical IoT applications are significant to consider. With erroneous priorities and high failure rates, catastrophic losses in terms of human lives, great business assets, and privacy leakage will occur in emergent scenarios. In this paper, an efficient resource slicing scheme for optimizing federated learning in software-defined IoT (SDIoT) is proposed. The decentralized support vector regression (SVR) based controllers predict the IoT slices via packet inspection data during peak hour central congestion to achieve a time-sensitive condition. In off-peak hour intervals, a centralized deep neural networks (DNN) model is used within computation-intensive aspects on fine-grained slicing and remodified decentralized controller outputs. With known slice and prioritization, federated learning communications iteratively process through the adjusted resources by virtual network functions forwarding graph (VNFFG) descriptor set up in software-defined networking (SDN) and network functions virtualization (NFV) enabled architecture. To demonstrate the theoretical approach, Mininet emulator was conducted to evaluate between reference and proposed schemes by capturing the key Quality of Service (QoS) performance metrics.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.5
• /
• pp.162-177
• /
• 2020

The perception of traffic environment based on various sensors in autonomous driving system has a direct relationship with driving safety. Recently, as the perception model based on deep neural network is used due to the development of machine learning/in-depth neural network technology, a the perception model training and high quality of a training dataset are required. However, there are several realistic difficulties to collect data on all situations that may occur in self-driving. The performance of the perception model may be deteriorated due to the difference between the overseas and domestic traffic environments, and data on bad weather where the sensors can not operate normally can not guarantee the qualitative part. Therefore, it is necessary to build a virtual road environment in the simulator rather than the actual road to collect the traning data. In this paper, a training dataset collection process is suggested by diversifying the weather, illumination, sensor position, type and counts of vehicles in the simulator environment that simulates the domestic road situation according to the domestic situation. In order to achieve better performance, the authors changed the domain of image to be closer to due diligence and diversified. And the performance evaluation was conducted on the test data collected in the actual road environment, and the performance was similar to that of the model learned only by the actual environmental data.