An open source artificial intelligence agent developed by researchers at Carnegie Mellon University (CMU) and Meta AI gives robots manipulation abilities equivalent to those of a three-year-old child through passive observations and active learning.
RoboAgent was able to complete 12 manipulation skills in differing real-world scenarios. The researchers gave the robot self-experiences from which it could learn by teleoperating it through various tasks. It also learned how humans interact with objects and leverage different skills to complete tasks from online videos.
"An agent capable of this sort of learning moves us closer to a general robot that can complete a variety of tasks in diverse unseen settings and continually evolve as it gathers more experiences," says Shubham Tulsiani, assistant professor at the CMU School of Computer Science's Robotics Institute. "RoboAgent can quickly train a robot using limited in-domain data while relying primarily on abundantly available free data from the Internet to learn a variety of tasks."
From Carnegie Mellon University
View Full Article
Abstracts Copyright © 2023 SmithBucklin, Washington, D.C., USA
No entries found