Jointly Learning Grounded Task Structures from Language Instruction and Visual Demonstration

33572

Knowledge Triples

Source Entity Relationship Target Entity
captures
appliedIn