Fuzzy Integral-Based Gaze Control of a Robotic Head for Human Robot Interaction

Area of Science:

• Robotics
• Artificial Intelligence
• Human-Robot Interaction (HRI)

Background:

• Developing human-like gaze control is crucial for natural human-robot interaction (HRI).
• Previous research often overlooked hardware implementation, real-time processing, and real-world environmental factors essential for practical HRI.

Purpose of the Study:

• To propose a fuzzy integral-based gaze control algorithm for robotic heads that operates in real-time within real environments.
• To enhance natural human-robot interaction by creating more human-like gaze behaviors.

Main Methods:

• Gaze control is formulated as a multicriteria decision-making problem using seven human gaze-inspired criteria.
• A fuzzy integral aggregates partial evaluations of candidate gaze directions based on user-defined preferences (fuzzy measures).
• Inhibition of return is applied to adjust global evaluation values for final gaze direction selection.

Main Results:

• The proposed algorithm was implemented and demonstrated on a robotic head.
• Effectiveness was validated through interaction and comparison scenarios in real-world settings.
• The algorithm successfully achieved real-time gaze control, considering multiple sensory inputs and user preferences.

Conclusions:

• The fuzzy integral-based gaze control algorithm offers a viable solution for achieving natural HRI.
• The approach effectively integrates multicriteria decision-making with fuzzy logic for real-time robotic gaze control.
• This work contributes to the advancement of intelligent robotic systems capable of more intuitive interactions.