Virtual Robot
Virtual robot research focuses on creating and utilizing simulated robots to advance robotics capabilities and reduce the cost and risk associated with physical robot development and testing. Current research emphasizes using virtual robots for training physical robots via reinforcement learning and imitation learning, often incorporating digital twins and large language models to improve adaptability, task planning, and human-robot interaction. This work is significant because it enables efficient development of complex robot behaviors, including manipulation, navigation, and human-robot collaboration, ultimately accelerating the deployment of more capable and robust robots in real-world applications.
Papers
July 16, 2024
March 19, 2024
June 16, 2023
March 5, 2023
February 10, 2023
June 18, 2022
June 14, 2022
May 13, 2022
May 7, 2022
April 7, 2022