Representing Skill Demonstrations for Adaptation and Transfer.

We address two domains of skill transfer problems encountered by an autonomous robot: within-domain adaptation and cross-domain transfer. Our aim is to provide skill representations which enable transfer in each problem classification. As such, we explore two approaches to skill representation which address each problem classification separately.  The first representation, based on mimicking, encodes the full demonstration  and  is  well  suited  for  within-domain  adaptation.   The second representation is based on imitation and serves to encode a set of key points along the trajectory, which represent the  goal  points  most  relevant  to  the  successful  completion of the skill.  This representation enables both within-domain and cross-domain transfer. A planner is then applied to these constraints, generating a domain-specific trajectory which addresses the transfer task.