SG-Bot is a novel object rearrangement framework that uses a coarse-to-fine scheme with a scene graph as the scene representation. It is lightweight, real-time, and user-controllable. The framework employs a three-fold procedure – observation, imagination, and execution. Initially, objects are discerned and extracted from a cluttered scene during the observation phase. These objects are then organized and depicted within a scene graph. This scene graph informs a generative model, which forms a fine-grained goal scene considering the shape information from the initial scene and object semantics. Finally, the initial and envisioned goal scenes are matched to formulate robotic action policies.
Publication date: 22 Sep 2023
Project Page: https://sites.google.com/view/sg-bot
Paper: https://arxiv.org/pdf/2309.12188
