Multi-Participant Interaction Role Classification and Action Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional human-robot interaction technologies are limited to one-to-one interactions, making it difficult to perform natural interactions when a robot interacts with two or more participants.
Innovation Solution
An interactive device and method that classify roles for each participant based on external stimulus signals, using a role classifying unit to determine participation degrees and action states, allowing for customized interaction operations tailored to each participant.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional one-to-one interaction technology is used, then the interaction control is simple, but the natural interaction with multiple participants cannot be achieved
Solution Approach 1:
The interaction control system is segmented into multiple independent role management modules, each handling a specific participant's interaction based on their assigned role (e.g., speaker, listener, observer). This segmentation allows the system to manage multiple participants without overwhelming complexity, as each module operates independently with standardized protocols.
Solution Approach 2:
A role assignment mechanism acts as an intermediary between the robot and multiple participants. The system assigns roles to participants based on their interaction behavior and the current conversation context, then uses these roles as intermediaries to control interaction flow. This mediator approach simplifies the robot's decision-making process by reducing complex multi-participant interactions to role-based protocols.
2Ease of operation
If role classification based on participation degree is implemented, then the interaction becomes more natural and responsive, but the processing time and computational load increase
Solution Approach 1:
The system performs preliminary role assignment at the beginning of interactions with multiple participants, establishing role relationships before the actual conversation begins. This preliminary action allows the robot to pre-configure interaction protocols for each role, reducing real-time processing requirements during the actual interaction and enabling more natural, responsive communication.
Solution Approach 2:
The role classification system continuously monitors interaction behavior and provides feedback to adjust role assignments dynamically. By using feedback mechanisms, the system can quickly adapt to changing interaction dynamics without requiring extensive computational analysis in real-time, as the feedback loop enables incremental role adjustments based on observed behavior patterns.
Data Source
AI summary
The present disclosure herein relates to an interaction device capable of performing an interaction with a human, and more particularly, to an interaction device capable of performing an interaction with a plurality of participants. The interaction device includes a role classifying unit configured to classify a role for each of a plurality of participants based on an external stimulus signal for each of the plurality of participants and an action adjusting unit configured to perform different interaction operations for each of the plurality of participants based on the role for each of the plurality of participants. An interaction device according to an embodiment of the present application classifies the roles of a plurality of participants according to a participation degree and/or an action state and provides a customized interaction operation for each of participants according to the classified roles. Therefore, it is possible to perform a natural interaction operation with a plurality of participants.