コロキアムB発表

日時: 12月4日(月)3限目(13:30-15:00)


会場: L2

司会: 垣内 正年
林 秀樹 M, 1回目発表 数理情報学 池田 和司, 笠原 正治, 久保 孝富, 日永田 智絵
title: Mathemathical Modeling for Mobility Management
abstract: In an era marked by an aging society, the local transport system faces challenges stemming from a lack of drivers. The shifting demographics have intensified the need for a comprehensive transportation policy that addresses the evolving dynamics of the workforce. The shortage of drivers, exacerbated by the demographic shift, has prompted a reevaluation of the existing activity model within the transportation sector. We are currently surveying the latest research and examining ways to solve local transport problems.
language of the presentation:Japanese
 
久睦 竜主 M, 1回目発表 大規模システム管理 笠原 正治, 藤川 和利, 原 崇徳
title: *** On Network Slicing in O-RAN Assisted by Machine Learning ***
abstract: *** In this presentation, we delve into the core significance of network slicing in the 5G landscape, tailoring wireless services to specific needs. Focusing on RAN slicing, we explore its crucial role in dynamic resource allocation within the Radio Access Network (RAN). Then, the spotlight turns to the RAN Intelligent Controller (RIC), a central orchestrator dynamically managing RAN slicing functionalities. Advancing in our discussion, we carefully examine how machine learning (ML) integrates into the RIC framework, enhancing its capabilities. Our attention centers on an illustrative paper that provides clear and practical insights into the efficient application of ML techniques for optimizing RAN slicing and network management, offering a comprehensive view of the evolving wireless network paradigm. ***
language of the presentation: *** English or Japanese (choose one) ***
 
高橋 顕太 M, 1回目発表 インタラクティブメディア設計学 加藤 博一, 和田 隆広, 神原 誠之, 藤本 雄一郎, 澤邊 太志
title: The Uncanny Feeling of Anime Face Dialogue Robots on Humans
abstract: In the field of human-agent interaction (HAI), there are various types of interactive robots ranging from those that resemble real humans to those that are mechanical in appearance. In Masahiro Mori's uncanny valley, it is stated that the more a robot's appearance resembles a human, the more its likability increases, but beyond a certain level of similarity, the likability decreases. In this study, the effectiveness of a dialogue robot with an anime face is examined with the aim of reducing the uncanny feeling one has toward robots. For this purpose, we will classify human-like and anime-like facial features along structural and material axes, and conduct experiments using virtual reality (VR) technology to simulate an interactive robot. Currently, we are investigating methods to parametrically generate intermediate faces that are distributed between human-like and anime-like faces.
language of the presentation: Japanese
 
新江田 航大 M, 1回目発表 インタラクティブメディア設計学 加藤 博一, 和田 隆広, 神原 誠之, 藤本 雄一郎, 澤邊 太志
title: Effect Evaluation of Pain Perception by Robot's Stroking with Speech
abstract: The purpose of this study is to verify whether the 'stroking with speech' behavior by the robot can change the perception of physical pain. People perform a 'stroking with speech' behavior that gives comfort to the other person during caregiving and nursing care. In recent years, robots that interact with humans have been attracting attention due to advances in artificial intelligence and robotics technology, and it has become known that 'stroking with speech' behavior by robots give people more comfortable feelings than 'speech' only or 'stroking' only. This study shows whether the 'stroking' and 'stroking with speech' behaviors are effective in reducing pain perception, and whether the 'stroking with speech' behavior, which has a greater psychological impact than the 'stroking' action, can reduce pain perception more based on the gate-control theory. In experiments, 37 subjects were evaluated, and the results were both hypotheses supported by the results.
language of the presentation: Japanese

 
渡辺 諒 M, 1回目発表 情報セキュリティ工学 林 優一, 安本 慶一, 藤川 和利, 藤本 大介, 鍛治 秀伍
title: A Study on Method for Detecting Recycled ICs via Intentional Electromagnetic Interference Focusing on Aging of Digital Output Circuits
abstract: Problems have been reported due to mounting counterfeit integrated circuits (ICs) in electronic devices. In particular, recycled ICs extracted from discarded electronic devices have similar functionality to fresh ICs, making them difficult to detect. A detection method for recycled ICs using backscattering side-channel analysis (BSCA) has been proposed. In a previous study, conducting BSCA on target circuits within ICs required precise probe alignment to their implementation positions. Therefore, it is difficult to apply the method to ICs where the implementing positions of the target circuits are unknown. This study investigates a detection method for recycled ICs against ICs in which the implement positions of the target circuit are unknown by conducting BSCA that focuses on the aging of the digital output circuits.
language of the presentation: Japanese
 
池内 陸 M, 1回目発表 情報セキュリティ工学 林 優一, 岡田 実, 向川 康博, 藤本 大介
title: A Study on Countermeasures Against Electromagnetic Information Leakage from Displays by Switching Masking Images
abstract: Devices with displays are used across various fields. The devices may display confidential information that causes economic losses if intercepted. The threats of reconstructing the display information by capturing the electromagnetic (EM) waves emitted from the display have been pointed out. The eavesdropper passively captures EM waves emitted from devices, making it difficult for users to detect the attack. One of the countermeasures is to process the image displayed on the device, such as using an image masked by noise. However, since the timing and order of the masked images are constant in the method, the display information may be leaked by extracting and reconstructing the output image of specific timing. This study demonstrates the feasibility of such attacks and investigates the countermeasures.
language of the presentation: Japanese