One-Shot Imitation Learning - Reimplementation of DOME paper

• Environment and dataset collection:

  • A total of 120 trajectory episodes were recorded using a Kinova arm in the ROS Gazebo simulation.

• Success Case - Zero-shot model:

• Success Case - One-shot model:

References:

• E. Valassakis, G. Papagiannis, N. Di Palo, and E. Johns, "Demonstrate Once, Imitate Immediately (DOME)," Proc. IEEE/RSJ IROS, 2022.