新任助教講演会(Lectures from New Assistant Professors)

日時(Datetime) 令和2年6月26日(金)3限 (13:30 -- 15:00), 2020/06/26, Friday
場所(Location) WebEx
司会(Chair) 小林 泰介 (Taisuke Kobayashi)

講演者(Presenter) 藤本雄一郎(Yuichiro Fujimoto), インタラクティブメディア設計学研究室 (Interactive Media Design Lab.)
題目(Title) Augmented Reality Based on the Estimation of the Human Internal State
概要(Abstract) In this presentation, I briefly overview two research fields I was working on: 1) Augmented reality (AR) for task support and 2) Estimation of human internal states. Augmented reality is the technology that overlays computer-generated information onto the real world. In the first field, I explored two possibilities of this technology to expand the applicability for human task support. In the second field, I developed the estimation algorithm of how busy the people are in the office or in his/her own home by using various sensor data (e.g., PC, smartphone, and body movement). Finally, the new research field overlapping of the above two themes is explained and I reveal that AR based on the estimation of human internal states enables us to change the appearance of the object to gives rise to the desired perception.

講演者(Presenter) 花田研太 (Kenta Hanada), 知能システム制御研究室 (Intelligent System Control Lab.)
題目(Title) Network Analysis via Artificial Intelligence: Boolean Satisfiability based Approach
概要(Abstract) Networks appear everywhere in modern society, for example, power systems, communications, and infectious disease models. Thus, it is significant to analyze the characteristics of networks and behaviors of systems on the network in order to realize better society. In this presentation, I will show my previous study of the network analysis by using Boolean satisfiability (SAT) based approach, which is one of the techniques of artificial intelligence (AI). This study focuses on cascading failures on power networks. A cascading failure is a phenomenon where one failure in a system causes another and this process continues until no further failures occur. Once a cascading failure has occurred in a power system, it brings massive blackouts and huge economic and social losses. A main question to be answered in the analysis of cascading failures is which nodes are vulnerable in the sense that their failures trigger system-wide failures. The proposed approach reduces the problem of analyzing cascading failures to the SAT problem by constructing Boolean formulas that symbolically represent all possible failure scenarios under given conditions. This allows us to examine a numerous number of scenarios by means of fast modern SAT solvers which have been studied and developed in AI research fields.