Imitation Learning with Dynamic Movement Primitives

Overview

Master’s Thesis at Boston University developing automation algorithms for reaching-to-grasping tasks using imitation learning.

Key Contributions

• Developed an automation algorithm for reaching-to-grasping task in V-REP simulation using imitation learning approach
• Leveraged joystick for the Baxter Robot end-effector control and human demonstration data collection
• Implemented learning from demonstration (LfD) algorithms for skill transfer