Theme-
6
公募
Publicly Offered Research
Publicly Offered Research
First term
Fully-automated makerless motion capture for group 3d motion analysis (Shohei Nobuhara)
This research is aimed at developing a new algorithm for markerless motion capture with a fully-automatic multi-view camera calibration. This contributes capturing 3D trajectories of group behavior automatically in the wild.
Elucidation of Behavioral Mechanisms of Biological Swarms by Mathematical Groups and its Application to Engineering Design (Kazunori Sakurama)
This research aims to develop a technique for constructing mathematical models to understand the emergence mechanisms of sophisticated swarming behaviors comprehensively. By creating such models from behavioral data, we will elucidate the emergence mechanisms of various swarming behaviors , including hierarchical navigation and decision-making, and develop engineering technology that mimics the functions of organisms.
Flexibility and robustness in the honeybee path-integration (Midori Sakura)
A single honeybee memorizes multiple foraging sites. How do these multiple memories interact each other in the decision-making process during navigation? By combining a trajectory analysis using a VR flight simulator and interruption of their navigation behavior by inducing a recall of another memory, we try to elucidate the navigational decision-making process operated by multiple place memories.
Understanding animal group navigation via nonlinear hierarchical representation learning (Hiroshi Morioka)
This study aims to understand the mathematical mechanism of navigation behavior of animal groups by proposing a novel hierarchical representation learning framework based on machine learning, which estimates a stochastic model of nonlinear hierarchical dynamics hidden behind time-series data observed from them in a data-driven manner.
Guide robot for human group inspired from animal navigation (Satoru Satake)
The purpose of this research is to develop social robots that guide a human group to destination. We apply the idea of animal navigation to reveal the mechanism of human group navigation that make robot navigate as a member of human group, and control the group if necessary. To evaluate the effeteness, we use the actual robot in a shopping-mall.
Second term
The Analysis of Social Relationships Construction in Songbirds and Computational Modeling of Social Navigation (Genta Toya)
The behavior of new individuals introduced into a group searching for the best social position can be modeled as a kind of navigation. I will analyze and formulate social behavior and relationship formation processes by quantifying the Bengalese finch's behavior in lab.
Elucidating the Hierarchical Inter-Agent Interaction Mechanisms for Realizing Robot Shepherding (Yusuke Tsunoda)
To realize the Robot-Shepherding System, we aim to develop a mobile robot: A shepherd-logbot, which indirectly stimulates a group of sheep from the outside. Through guidance experiments with a flock of sheep, we will elucidate the response system of the actual sheep and the interaction mechanism between the mobile robot and the sheep.
Development of a biologging system for measuring airflow speed during seabird migration (Hidetoshi Takahashi)
Large seabirds are known to engage in dynamic soaring, in which they glide in a meandering manner and use the vertical gradient of crosswinds on the sea surface. This research aims to construct a measurement system for airflow speed of such seabirds in flight, using a pitot-tube type sensor that meets strict specifications such as waterproof, small size, light weight, and low power consumption.
Machine Learning-Based Modeling for Understanding Hierarchical Navigation in Diverse Animal Groups (Kazushi Tsutsui)
This research aims to elucidate the relationship between group dynamics and individual behavioral characteristics by using multi-agent computational simulations based on hybrid models that integrate deep learning and biological theory, and the relationship between individual behavioral characteristics and the neural bases by collaborating with the Physical Team.
The Theory of Why We Get Lost: Behavioral Changing by Intervening in the EEG during playing VR Maze (Arao Funase)
This research focuses on "getting lost" when humans move toward a destination.In particular, we will attempt to identify the EEG involved in decision-making about the direction of locomotion at the branches of a maze, and to elucidate "getting lost". In addition, we will conduct an intervention experiment to clarify whether spatial memory can be controlled.
Dynamical Analysis of Biological Individuals using Inertial Sensors (Tsuyoshi Mizuguchi)
We focus on the motion of biological individuals and aim to clarify their dynamics through kinematic analysis of measured data. We mainly treat the time series of the acceleration and angular velocity obtained by inertial sensors attached to each individual, and grasp their spatio-temporal behavior from the multiple view points such as gaits, energetics, control and collective effects
-
A02-6 公募 研究者
-
Principal Investigator
Shohei Nobuhara
Shohei Nobuhara
Faculty of Information and Human Science, Kyoto Institute of Technology
Professor
- Research field
- Computer Vision
- Research keywords
- 3D shape and motion estimation, multiple-view geometry, camera calibration, physics-based vision, computational photography
- Laboratory
- https://www.is.kit.ac.jp/research.html
-
Principal Investigator
Kazunori Sakurama
Kazunori Sakurama
Graduate School of Informatics, Kyoto University
Associate Professor
- Research field
- Control Engineering
- Research keywords
- Distributed Control, Distributed Optimization, Swarm Robotics, Formation Control, Group Theory
- Laboratory
- http://www.ids.sys.i.kyoto-u.ac.jp/sakurama/index_e.html
-
Principal Investigator
Midori Sakura
Midori Sakura
Graduate School of Science, Kobe University
Associate Professor
- Research field
- Neuroethology, Neurobiology
- Research keywords
- Insects, Path-integration, Flight simulator, VR, celestial compass
- Laboratory
- .
-
Principal Investigator
Hiroshi Morioka
Hiroshi Morioka
RIKEN Center for Advanced Intelligence Project (AIP)
Research Fellow
- Research field
- Machine learning, Computational neuroscience
- Research keywords
- Machine learning, Nonlinear independent component analysis, Unsupervised representation learning，Deep learning, Brain-imaging
- Laboratory
- https://sites.google.com/view/hiroshimorioka
-
Principal Investigator
Satoru Satake
Satoru Satake
Deep Interaction Laboratory Group, Advanced Telecommunications Research Institute International
Researcher
- Research field
- Human robot interaction, Social Robot
- Research keywords
- Human-robot interaction, Social robot, Social navigation, Group navigation, Sensor network
- Laboratory
- https://dil.atr.jp/ISL//en/
-
Principal Investigator
Genta Toya
Genta Toya
Research Center for Advanced Science and Technology, the University of Tokyo
Project Assisitant Professor
- Research field
- Complex Systems, Knowledge Science
- Research keywords
- Agent-based modeling, evolutionary simulation, behavior tracking, niche construction, alienation
- Laboratory
- https://www.researchgate.net/profile/Genta-Toya
-
Principal Investigator
Yusuke Tsunoda
Yusuke Tsunoda
Graduate School of Engineering, University of Hyogo
Assistant Professor
- Research field
- Control engineering, swarm robot system, swarm system, disaster response robot, rough terrain mobile robot
- Research keywords
- Sheepdog system, collective model, robot shepherd, system identification, control engineering
- Laboratory
- .
-
Principal Investigator
Hidetoshi Takahashi
Hidetoshi Takahashi
Faculty of Science and Technology, Keio University
Associate Professor
- Research field
- Mechanical Engineering
- Research keywords
- Force sensor, MEMS, Biomechanics, Mechatronics, Precision machining
- Laboratory
- http://www.takahashi.mech.keio.ac.jp/
-
Principal Investigator
Kazushi Tsutsui
Kazushi Tsutsui
Graduate School of Arts and Sciences, The University of Tokyo
Assistant Professor
- Research field
- Machine learning, Behavioral science
- Research keywords
- Multi-agent simulation, reinforcement learning, collective behavior, cooperation, computational biology
- Laboratory
- .
-
Principal Investigator
Arao Funase
Arao Funase
-
Principal Investigator
Tsuyoshi Mizuguchi
Tsuyoshi Mizuguchi
Graduate School of Science, Osaka Metropolitan University
Associate Professor
- Research field
- Nonlinear Physics
- Research keywords
- Inertial sensor, kinematic analysis, gaits, energetics, collective behavior
- Laboratory
- https://www.omu.ac.jp/sci/nonlinear/
