Situated Bayesian Reasoning Framework for Robots Operating in Diverse Everyday Environments

摘要

When a robot is tasked to operate in a new environment, it should have the ability to leverage external knowledge sources to acquire common knowledge about its general environment instead of learning everything from scratch. For example, a maintenance robot should have the ability to leverage common knowledge about tools, just as a home robot should have access to knowledge about household items. Multiple knowledge bases and semantic knowledge graphs have been developed in the AI community that incorporate general, commonsense knowledge; examples include WordNet, ConceptNet, and ResearchCyc, as well as extensions of this work in other research communities, such as ImageNet in computer vision. Prior work has shown that these language-based knowledge resources can be used as a foundation for powerful reasoning methods.