This AI paper proposes LLM-Grounder: a zero-shot, open-vocabulary approach to 3D visual grounding for next-generation home robots
Understanding their environment in three dimensions (3D vision) is essential for home robots to perform tasks such as navigation, manipulation ...