| 近江谷 旦 | 1751020: M, 1回目発表 | サイバーレジリエンス構成学 | 門林 雄基 |
|
title: Educational Tools to Learn Concrete Control Measures Corresponding to Threats of Cyber Attacks
bstract: There is the need to understand information security measures to be implemented in order to effectively defend its own system and protect the assets from cyber attacks. We have developed an educational tool, card game from which we can learn measures in order to efficiently protect our own systems from various existing attack methods. This tool imports gamification approach that has been confirmed its effectiveness in maintaining and improving our enthusiasm for learning. language of the presentation: Japanese | |||
| 白石 裕輝 | 1751055: M, 1回目発表 | サイバーレジリエンス構成学 | 門林 雄基 |
|
title: A Locality-Aware Object Store in Edge Computing Environments abstract: Novel computing paradigms such as Multi-access Edge Computing (MEC) and Fog Computing are proposed in order to support the Internet of Things (IoT) service requirements. These paradigms, called edge computing, extend a part of cloud computing resources to network edges and provide ultra-low latency services to end users. There is a huge demand for object stores since latency-sensitive IoT services require fast and scalable storages. In this work, we review the overview of the edge computing and discuss the existing object store. We also propose a locality-aware object store that reduces interactions with other edge nodes. language of the presentation: Japanese | |||
| 佐々木 俊樹 | 1751048: M, 1回目発表 | 自然言語処理学 | 松本 裕治 |
|
title: Data Augmentation for Reading Comprehension Task
abstract: A major goal for NLP is for machines to be able to understand text as well as people. Such an ability can be tested with reading comprehension(RC) task that requires the agent to read open-domain documents and answer questions about them. I introduce data augmentation methods for low-resource RC questions. language of the presentation: Japanese | |||
| 桶田 真吾 | 1751024: M, 1回目発表 | サイバネティクス・リアリティ工学 | 清川 清 |
|
title: A study on transition between real and VR spaces via virtualized reality environments by omnidirectional video
abstract: Users in conventional virtual reality (VR) systems with a head mounted display (HMD) usually do not perceive and recognize their experiences as real. For example, the learning effect in VR applications such as a disaster simulation decreases because users recognize their experiences as a fiction. In this study, we propose a new real - VR transitional interface in which users can keep their belief in the reality while experiencing a VR system. The proposed interface smoothly changes the user’s view from the real space to a VR space temporally and spatially by way of an omnidirectional video captured at the user position. In this presentation, we report the proposed interface and a prototype application to experience a dwarf's vision that we created as an example. language of the presentation: Japanese | |||
| 古賀 隆文 | 1751041: M, 1回目発表 | サイバネティクス・リアリティ工学 | 清川 清 |
|
title:
Image completion of peripheral regions of existing video
contents for enhancing immersion
abstract: In this study, we present an image completion method that fills in peripheral regions outside a video frame to enhance immersion of existing video contents presented with immersive display systems.The proposed method employs generative adversarial networks (GANs). Assuming that regions near the border of an input image have high correlations to their outer regions, we design dedicated GANs that divides the input image into a grid and applies a pair of generator and discriminator networks to each grid.While each discriminator keeps local consistency, we additionally employ a discriminator for the entire image to keep global consistency.We will evaluate the visual plausibility of the completed images with a user study. language of the presentation: Japanese | |||
| 寺田 周平 | 1751074: M, 1回目発表 | サイバネティクス・リアリティ工学 | 清川 清 |
|
title:Designing of an overeating suppression system by visual and haptic feedback based on food intake
abstract:One of the values we pursue in a meal is the sense of satiety. However, continuing pursuit of satiety with daily meals may result in loss of control of food intake and health problems. Therefore, in this study, we propose an abdomen-wearing device that presents visual and haptic stimulus in order to reduce food intake while pursuing the sense of satiety. By sensing the amount of food intake by the user and presenting visual and haptic feedback at an appropriate timing according to the result, the system will make the user feel he or she has eaten too much when it actually is a proper amount. In this way, our system improves the user's self-control and aids suppression of food intake. In this presentation, we describe an overview of the proposed system and the prototype under development. language of the presentation: Japanese | |||
| 渡邊 貴弘 | 1751131: M, 1回目発表 | サイバネティクス・リアリティ工学 | 清川 清 |
|
title:
An AR badminton support system that adjusts the difference between
skill levels of novice and experienced players
abstract: Badminton is a popular sport both as competition and a lifelong activity. However, due to the high deceleration characteristics of the shuttlecock, it is difficult to predict the timing of the swing, and it causes a large difference in skill level between novice and the experienced players. As a result, it is necessary for experienced players to adjust to the level of novice players, which hampered the exhilarating sports experience. Therefore, in this study, we will develop an AR badminton support system that adjusts the difference in skill between them. As a support for novice players, our system predicts a future trajectory of the shuttlecock based on the deceleration characteristics and displays it onto the user's optical see-through HMD. At the same time, it also calculates and displays an effective shuttlecock hit point on the trajectory, which is expected to be offensive or easy to swing. We plan to verify that the successful receive ratio improves with the support system and that the shot angle of the shuttlecock becomes steeper. language of the presentation: Japanese | |||
| 蓼沼 知秀 | 1751068: M, 1回目発表 | 知能システム制御 | 杉本 謙二 |
|
title: The design of sampled-data observer with aperiodic sampling using switched gain
abstract: In sampled-data controlling systems, there could occur transmission loss when using a contactless measuring instrument or network communication channels. It makes sampling rates aperiodic, which causes the system problems such as estimation values of system's states diverge with a periodic design of observer. To prevent this problem,I propose the design of aperiodic sampling observer switching its gain depending on the elapsed time since the last sampling of measurement signal. I also execute the simulation of state estimation using proposed observer and conventional one having fixed gain and find that proposed one works successfully and performs better than the conventional one. language of the presentation: Japanese | |||
| 原田 潤一 | 1751096: M, 1回目発表 | 知能システム制御 | 杉本 謙二 |
|
title: Optimal Resource Allocation in Product Development Processes
abstract: We study dynamical processes of product developments from the viewpoint of systems and control theory. We first model the product development process as a deterministic and positive switched linear system. We then formulate our resource allocation problem for efficient product development as an optimization problem on the final state of the switched system. We show that the optimization problem can be converted to a convex program. We numerically confirm our theoretical results using a practical example in the context of vehicle productions. language of the presentation: Japanese | |||
| 岡西 孝真 | 1751023: M, 1回目発表 | 計算システムズ生物学 | 金谷 重彦 |
|
title: non-negative matrix factorization considering gene network structure
abstract: There are estimated 20,000~30.000 human genes. Because a phenotype is formed by many genes getting mixed up, gene network analysis has been performed to identify relationships between genes. However, the analysis result is complex and hard to understand. Therefore, in this research, We propose non-negative factorization considering gene network structure to help in understand of network analysis result. language of the presentation: Japanese | |||
| 齋藤 俊 | 1751043: M, 1回目発表 | 計算システムズ生物学 | 金谷 重彦 |
|
title: Drug efficacy prediction based on molecular graph topology abstract: Machine learning approaches have been widely applied to the field of drug discovery over the past decades. This research takes advantage of molecular geometrical information by representing compounds as undirected graphs. In this study, using data sets of compounds in medicine, we demonstrate the effectiveness of the graph-based method for prediction of drug efficacy. language of the presentation: Japanese | |||
| 武末 祥太郎 | 1751067: M, 1回目発表 | 計算システムズ生物学 | 金谷 重彦 |
|
title: The interpolation method of herb's metabolites data for predicting efficacy of traditional medicines from Indonesia
abstract: Indonesia has the largest medicinal plant species in the world and these plants are used as Jamu medicines. Jamu medicines are popular traditional medicines from Indonesia and I need to systemize the formulation of Jamu and develop basic scientic principles of Jamu. However, the analysis of the metabolites in Indonesian herb plant are not carry out. Therefore, in order to solve these problems, I propose data interpolation method by using collaborative filtering to make up the information of herbal metabolites. language of the presentation: Japanese | |||
| 田中 徹士 | 1751069: M, 1回目発表 | 計算システムズ生物学 | 金谷 重彦 |
| title: Analysis of Soil Microbiome abstract: In soil, a wide variety of microbes exist, and these form communities called microbiome. I analyzed soil microbiome taken from a variety of countries and environment and found 3 independent ecosystems. Each of ecosystem has dominant microbe and they are independent with geological features. | |||