Bootstrapping the Semantics of Tools: Affordance Analysis of Real World Objects on a Per-part Basis

This study shows how understanding of object functionality arises by analyzing objects at the level of their parts where we focus here on primary tools. First, we create a set of primary tool functionalities, which we speculate is related to the possible functions of the human hand. The function of a tool is found by comparing it to this set. For this, the unknown tool is segmented, using a data-driven method, into its parts and evaluated using the geometrical part constellations against the training set. We demonstrate that various tools and even uncommon tool-versions can be recognized. The system "understands" that objects can be used as makeshift replacements. For example, a helmet or a hollow skull can be used to transport water. Our system supersedes state-of-the-art recognition algorithms in recognition and generalization performance. To support the conjecture of a possible cognitive hand-to-tool transfer we analyze, at the end of this study, primary tools by also incorporating tool-dynamics. We create an ontology of tool functions where we find only 32 of them. Being such a small set this would indeed allow bootstrapping tool-understanding by exploration-based learning of hand function and hand-to-tool transfer.