Leveraging Large Language Models for Robot 3D Scene Understanding

利用LLMS进行场景理解的工作
代码：https://github.com/MIT-SPARK/llm_scene_understanding

leveraging language for classifying rooms in indoor environments based on their contained objects:
(i) a zero-shot approach,
(ii) a feed-forward classifier approach, and
(iii) a contrastive classifier approach.

本质是利用语言模型学习到的’common-sense’机制来进行类别划分，同时对发现新类和未知类有益。

Methods: summarize a room’s contents in a query sentence