Hierarchical Task-Parameterized Learning from Demonstration for Collaborative Object Movement

Figure 5

Illustration of Equations (2), (3) and (4) and the morphing functions in Algorithm 3. The original task frames are randomly sampled, and the original trajectory and GMM clusters are generated from the demonstration strategy and trained accordingly. The morphed frames, trajectory, and GMM clusters are calculated from the original ones with . Note that the original and morphed start frames coincide.