Body-Oriented Gesture Generation System for Medical Interpreter Robots Based on Reinforcement Learning from Human Feedback

Medical interpreters are crucial in facilitating communication
between healthcare stakeholders who speak different languages. Bodyoriented
gestures convey critical information essential for accurate and
high-quality interpretation in healthcare settings. This study introduces
a body-oriented gesture generation system designed for medical interpreter
robots based on reinforcement learning from human feedback
(RLHF). The system allows robots to interpret more naturally and improve
over time through interactions with humans. By adopting a humancentered
development approach, we tailor our system to address the actual
needs of healthcare stakeholders.