A Visual Analogy Approach to Source Case Retrieval in Robot Learning from Observation

Abstract

Learning  by  observation  is  an  important  goal  in  developing complete intelligent robots that learn interactively.   We present a visual analogy approach toward an integrated, intelligent system capable of learning skills from observation. In particular, we focus on the task of retrieving a previously acquired case similar to a new, observed skill.  We describe three approaches to case retrieval:  feature matching, feature transformation, and fractal analogy. SIFT features and fractal encoding were used to represent the visual state prior to the skill demonstration, the final state after the skill has been executed, and the visual transformation between the two states. We discovered that the three methods (feature matching, feature  transformation,  and  fractal  analogy)  are  useful  for  retrieval of similar skill cases under different conditions pertaining to the observed skills.

A Visual Analogy Approach to Source Case Retrieval in Robot Learning from Observation

Posted in .