Plan-based Object Search and Exploration Using Semantic Spatial Knowledge in the Real World

• In this paper we present a principled planner based approach to the active visual object search problem in unknown environments. We make use of a hierarchical planner that combines the strength of decision theory and heuristics. Furthermore, our object search approach leverages on the conceptual spatial knowledge in the form of object cooccurences and semantic place categorisation. A hierarchical model for representing object locations is presented with which the planner is able to perform indirect search. Finally we present real world experiments to show the feasibility of the approach.

Subject headings

NATURVETENSKAP  -- Data- och informationsvetenskap (hsv//swe)

NATURAL SCIENCES  -- Computer and Information Sciences (hsv//eng)

Keyword

Active Sensing

Object Search

Semantic Mapping

Publication and Content Type

ref (subject category)

kon (subject category)