Wang Xiao (王啸)Ph.D. Student
Department of Computer Science and Engineering |
 |
Revolutionizing Therapy with LLM-Robot Agents
Abstract
This project integrates large language and vision-language models into Misty robots, enabling deep contextual understanding and advanced human-robot interactions. By implementing real-time pose estimation and emotion detection from video streams, the system dynamically adjusts Misty’s interaction strategies, ensuring more responsive and tailored user experiences. In the educational sphere, Misty serves as a platform for large-scale models to provide personalized guidance for children with special needs. Through real-time analysis of posture, expressions, and behavior, Misty refines instructional complexity, conversational content, and engagement methods, fostering greater interest, participation, and adaptability in therapeutic and learning interactions.
|
Last Updated on March, 2024
|
