Human Symbiotic Systems LaboratoryKatagami Lab. / Faculty of Engineering, Tokyo Polytechnic University

RESEARCH

Research

Exploring relationships between people and AI from multiple perspectives.

HOME / Research

Driving support agents

Driving simulator used in driving support agent experiments
Research environment using a driving simulator
  • Developing agents to build new relationships between cars and drivers
  • Creating a passenger effect through dialogue with an agent while driving
  • Driving-status notification agents that build passenger trust during the introduction of fully automated vehicles
  • Effects of driving support agents on hazard anticipation
  • Improving drivers' safety-checking behavior with a coaching agent using head-orientation estimation
  • Relationships between driver characteristics and overtrust in driving support robots
  • Developing a companion agent that creates a passenger effect through casual conversation
  • Developing an embodied agent to reduce drivers' mental workload
  • Interaction design using on-board diagnostics (OBD) to build new relationships between cars and drivers

Dialogue systems

Research on conversation with a robot in a smart home
Dialogue and interaction in everyday living spaces
  • Developing an affinity-building dialogue system using speech-level shifts based on DP theory
  • Developing an affinity-building dialogue system based on politeness theory and natural conversation corpora
  • A dialogue system controlling sentence endings according to speaker attributes
  • Effects of compliments by conversational agents
  • Developing a game-master bot for tabletop role-playing games and analyzing the effects of ambiguous utterances
  • Developing a job interview support system using multimodal information
  • Effects of an embodied agent's persuasiveness when recommending novels
  • Developing an automated esports commentary system
  • Developing a dementia care communication training system using embodied agents

Human–Agent Interaction (HAI)

Research illustration of a person talking with a robot
Exploring relationships between people and agents — research illustration
  • Developing educational materials through cross-cultural simulation games
  • Animacy for Anything: a device that adds lifelike expressions
  • Developing a dining companion agent
  • Developing an announcer agent that provides information
  • Developing interactive digital signage
  • Developing discommunication robots

AIWolf

AIWolf presentation at CEDEC2017
AIWolf research presentation / CEDEC2017
  • Developing AIWolf agents
  • Persuading human players with AIWolf
  • Developing an agent that explains the situation and helps organize information in Werewolf games
  • Natural utterances for AIWolf agents
  • Persuasive utterances for AIWolf agents

Mood Engineering

Conceptual illustration of the atmosphere between people and agents
Interactions that create emotion, presence, and trust — conceptual illustration
  • Improving trust in embodied agents
  • Creating social atmospheres with embodied agents
  • Creating an artificial atmosphere conveying familiar presence through Twitter and a social conversational robot
  • Assessing sociality with an agent using egograms

Information visualization

Conceptual network diagram showing connections between research fields and information
Making connections between information visible — conceptual illustration
  • Using social networks to support people stranded after earthquakes
  • Developing a system to prevent online backlash on Twitter
  • Assessing and visualizing author activity using publication submission data
  • Revealing the hidden specialties of journals through hierarchical research-field networks
  • Understanding hidden specialties through research-field networks
  • Analyzing networks between research fields using bibliographic information