• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• MISULJARYO - National Museum of Korea Art Journal
• /
• v.100
• /
• pp.112-138
• /
• 2021

This paper mainly analyzes the records titled 'Shijujil(施主秩)' from the Bokjangs of each of the Rocana and Shakyamuni statues enshrined as wooden seated Vairocana Buddha Triadcomposed of Vairocana(center), Rocana(right), and Shakyamuni(left) at the Daeungjeon Hall of Hwaeomsa Temple in Gurye. The Shijujil from the Shakyamuni statue was recovered through Bokjang investigation in September 2015 and has been kept in the museum of Hwaeomsa as an undisclosed relic. After the discovery of the Shijujil from the Rocana statue through an Bokjang investigation in July 2020, both of the Shijujils were only officially released through the special exhibition 'Grand Hwaeomsa Temple in Jirisan Mountain' in September 2021. Existing documents recording on the creation of Buddha statues in the 17th century are in the form of sheets or rolls. However, the Shijujils take the form of simple stitched booklets. The Shijujil from Rocana consists of 19 chapters and 38 pages in one book, and the Shijujil from Shakyamuni consists of 11 chapters and 22 pages in one book. The contents of the Shijujils consist of the purpose of the Buddha statue creation, the creation date, the year and place of enshrining, the names of the statues, the people in charge and their roles, the sculptors, the list of items donated, and the list of the contributors. In addition, the list of monks who were staying at Hwaeomsa Temple at that time are also recorded, so the Shijujil is like a time capsule that tells the situation of Hwaeomsa Temple about 400 years ago. According to the records of the Shijujils and the Writing on the wooden pedestal of Rocana, the Vairocana Triad began to be in March 1634(12th year of King Injo) and was completed in August of that year, and was enshrined in the Daeungjeon Hall in the fall of the following year. It is very important to confirm that the Vairocana Buddha Triad of Hwaeomsa was created in 1634. Since studies on the reconstruction of Hwaeomsa Temple in the 17th century and the roles of Byeokam Gakseong have been mainly based on 『湖南道求禮縣智異山大華嚴寺事蹟』 written by monk Haean in 1636, it has been estimated that the wooden seated Vairocana Buddha Triad was created in 1636. However, it is now known that the Virocana Buddha Triad was created in 1634. The Shijujils are also a good source of information about Byeokam Gakseong who played a pivotal roles in the reconstruction projects of Hwaeomsa Temple in the 17th century. He played leading roles in rebuilding the East Five-story Stone Pagoda(1630), in creating the wooden seated Vairocana Buddha Triad(1634), and in producing the Yeongsanhoe Gwaebul(1653, Hanging Scroll Painting depicting the Shakyamuni preaching). It is also very important that the Shijujils are records that can reveal the relationship between Byeokam Gakseong and royal family of Joseon Dynasty in the 17th century. The Shijujils from Rocana and Shakyamuni are the first documents ever discovered in which the names of royal family members, such as Uichanggun(Gwang Lee, son of King Seonjo), Ikseong Shin(son-in-law of King Seonjo), and Crown Prince Sohyeon(son of King Injo) are recorded in detail in relation to the production of Buddha statues. The Shijujils from Rocana and Shakyamuni contain specific information about the production of the wooden seated Vairocana Buddha Triad in the 17th century, such as the year of production of the Buddha statues, the role of Byeokam Gakseong, and the relationship between Byeokam Gakseong and the royal family, so it is of great value not only for art history but also for historical studies of Hwaeomsa Temple.

• Korean Journal of Materials Research
• /
• v.2 no.4
• /
• pp.270-278
• /
• 1992

The effect of Ni-rich phase on the stress-rupture properties of Ni mlcroalloyed W were studied using direct load creep tester at 100$0^{\circ}C$, 110$0^{\circ}C$, and 120$0^{\circ}C$ in $H_2.$ The stress rupture strength of 100hrs. of W-0.4wt% Ni was 43% higher at 100$0^{\circ}C$ and 35% higher at 110 $0^{\circ}C$than that of W-0.2wt% Ni due to the larger initial grain size, the higher relative density and the higher grain growth during test. That of W-0.8wt% Ni was 90% higher at 100$0^{\circ}C$ and 60% higher at 110$0^{\circ}C$ than that of W-0.2wt% Ni. The activation energy of W-0.4wt% Ni for creep was 81.3 Kcal/mole. It was considered that creep deformation was controlled by the diffusion of W in the Ni rich phase between the grains and the deformation of grains. All of the specimens showed intergranular fracture by grain boundary cavitation and growth of cavity throughout the entire spcimen cross-section.

• Journal of KIISE:Information Networking
• /
• v.33 no.3
• /
• pp.277-287
• /
• 2006

Most of the existing broadband access networks based on DSL, cable modem, and Ethernet support the best-effort internet access service, and adopt the flat rate pricing mechanism. It is almost impossible to provide the differentiated communication services, in current broadband access networks, for the different users and/or the different application services. Currently, however, the advances in multimedia, communication, and security technologies push the interactive and/or streaming multimedia services and VPN services to be widely deployed over Internet, and they require more QoS-sensitive services than the best-effort service. Though various QoS technologies such as RSVP-based IntServ and DiffSern were already developed and under standardization in Internet world, it is impractical to replace the existing QoS-unaware access networks with the QoS-enabled ones at a time to deploy QoS-sensitive services. In this paper, after analyzing current broadband access network architectures and the status of QoS support, we propose a practical approach to support multimedia QoS in the broadband access networks. The approach will be based on the integration of the differentiated pricing and the DiffServ technology. And it will be a step-wise approach to support backward compatibility with the legacy broadband access networks as much as possible.

• Journal of KIISE:Computing Practices and Letters
• /
• v.12 no.2
• /
• pp.91-101
• /
• 2006

Researches on services of vehicles have been mainly focused on how to provide useful information and entertainment for an in-vehicle driver. However, the needs are appreciably increased for more advanced services that help drivers to check and manage their vehicles conveniently, without requiring drivers to attach to their vehicles. It is a sort of ubiquitous computing, providing an intelligent interactive services for human at any time and any where. In this paper, we present an intelligent vehicle information system to enable a driver to remotely diagnose and control a vehicle via CDMA communication network connected to the Internet. The system improves mobility for diagnosis and control of vehicle by implementing a cut and call back mechanism, which allows the vehicle terminal to have access to the information server on the Internet via CDMA call. No matter where the driver is, he can obtain the remote diagnosis and control services on the web browser without any additional application installation. Design methodology is introduced and evaluation results are analyzed for the CDMA-based intelligent vehicle information system. The experimental results show that the response time of the vehicle terminal to a web client request is 10.302 seconds at the beginning and 646.44ms thereafter. The average response time of CAN sensor node to a vehicle terminal request is 6.669ms.

• Korean Journal of Environment and Ecology
• /
• v.26 no.3
• /
• pp.454-462
• /
• 2012

This study was to research the effective use of green rooftop space, facilities and gardening, targeting members from community rehabilitation centers with disabilities. The three community rehabilitation centers studied were, Namyangju Center located in a rural area, Seoul Center located in a urban area and Siheung Center located in both a rural and urban area. We analyzed the difference in preference on the basis of each local community area. In fact, it indicated that 50% of each center knew about the green rooftop at their facilities and its use as a place for taking walks and conversation. It also showed that there was the high preference for priority objects such as a bench, pergola and trash can. Also the preference for natural visualizations like herbal or ornamental plants. The study showed a high preference to a small vegetable plot, hands on gardening and ecological wetland. It also indicated that there was a high preference for experience in nature programs on the rooftops (28.9 %) versus the rate of horticultural programs (27%). Therefore, it proves that the composition of a green rooftop at a community rehabilitation center should be differentiated so that the green rooftop can be a place not only for resting, but also great for a natural learning experience and gardening therapy for people with disabilities.

• Korean journal of communication and information
• /
• v.61
• /
• pp.5-27
• /
• 2013

Media coverage of depression, the mental disorder, is on the high rise following the soaring number of reported celebrity suicide. This study is an exploratory attempt to get a glance on how online news media are portraying depression. The content analysis results indicate that celebrity was the most cited source, outnumbering the others such as non-celebrity patients and experts. More than half of the sample attributed the cause of depression to socio-psychological factors. Medical consultation was the most reported means of treating depression among the sample, while over the half did not suggest any treatment methods at all. Overall, celebrity related news were less likely to talk about the cause and treatment methods. In addition, the more neuro-biological factors were designated as the main cause of depression in the articles, the more chances of treatment method of all kinds were brought up. The frame of human interest dominated a little less than half of the articles examined, and only few reported positive outcome or achievements after coping with depression.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.43 no.3
• /
• pp.92-100
• /
• 2015

Unlike most other planning processes, brownfield planning generally requires a high level of technical and legal expertise due to potential site contamination. To successfully engage in inclusionary decision making, an adaptive collaboration strategy for brownfield planning is therefore critical. This study examines how a communicative planning approach can be used to overcome the challenge of enabling experts from different fields to work alongside lay people from the local community to achieve a properly balanced collaboration in brownfield planning. After identifying appropriate indicators for collaboration through a literature review of established communicative planning theory, these indicators are applied to the brownfield planning process, highlighting critical points of collaboration such as site prioritization, assessment, remediation, and redevelopment throughout. The results suggest the critical need for an adaptive model focusing on three aspects: 1. Facilitation of a balanced dialogue between the experts with social, cultural, and design-based knowledge and the ones with scientific and engineering-based knowledge, 2. Preparation of an appropriate tool for risk communication with the lay people, 3. Development of decision support system for the integration of expert-oriented technical data and public opinion-oriented subjective data.

• Korean Journal of Childcare and Education
• /
• v.11 no.2
• /
• pp.331-354
• /
• 2015

This study attempts to reveal the difficulties of teachers during the process of infant's early adaptation in the day care center and realize the thought and demand of teachers. And the purpose of this study is to make a desirable development in infant's early adaptation. The participants of the study were 9 infants all 1 year of age, their 7 parents, and 8 teachers of S public child day care center in P city. The 8 teachers were interviewed several times. The results of the study can be summarized as follows. First, the teachers' difficulties in their relationships with the parents are 1) egotism focusing on only their own child, 2) distrust between the teachers and parents, 3) a lack of understanding in the role as parents at early adaption programs, 4) and a lack of understanding in the management of the day care center. Second, in their relationships with the infants, the teachers'difficulties are 1) the uncontrollable behaviors of infants, 2) the arbitration of struggle between children, 3) and the inequality between enrolled and fresh infants. Third, in their relationships with the teachers, the difficulties are 1) co-work with other teachers 2) and conversation with the director of the day care center. This study is expected to give actual help to understand the difficulties of teachers who experience difficulties in the early infant adaptation process and to provide important preliminary data to plan specific management which assist with the difficulties of teachers.

• Cartoon and Animation Studies
• /
• s.33
• /
• pp.363-382
• /
• 2013

Up until about ten years ago, the use of typography played only an auxiliary role on broadcast television programs, primarily by transmitting information in order to facilitate a basic understanding of content. Recently, however, kinetic typography has become an important component in broadcast production. In fact, kinetic typography has developed into a visual language and a means of artistic expression, one that is increasingly used in the production of entertainment programs on television. This paper analyzes six aspects of kinetic typography: manner of development, location, intent, expressive techniques, color and font selection. Particular attention is placed on their use in three highly rated television entertainment programs: "2 Days & 1 Night", "Running Man", and "Infinite Challenge". The development way consists of the technique : starts off with cut and ends with cut. While, other techniques show conversation and situation representation using Z axis : zoom-in, zoom-out in , X axis : pan in <2 Days & 1 Night>. and Y axis : tilt in . Typographic design elements, expression technique, color, font are shown up according to the feature of each program. The resulting analysis suggests new ways for motion arts designers and the broadcast media to use kinetic typography in the development of television programs.