Grounded Lexicon Acquisition - Case Studies in Spatial Language
This work addresses the challenge of grounded language acquisition for robots, which is incremental as it builds on existing methods in spatial language learning.
The paper tackles the problem of enabling humanoid robots to acquire English spatial lexicons from robot tutors, identifying how systems like projective, absolute, and proximal spatial language can be learned without relying on direct meaning transfer or access to interlocutors' world models, and demonstrating that multiple systems can be acquired simultaneously.
This paper discusses grounded acquisition experiments of increasing complexity. Humanoid robots acquire English spatial lexicons from robot tutors. We identify how various spatial language systems, such as projective, absolute and proximal can be learned. The proposed learning mechanisms do not rely on direct meaning transfer or direct access to world models of interlocutors. Finally, we show how multiple systems can be acquired at the same time.