Embodied Scene Rearrangement Planning

Embodied Scene Rearrangement Planning

|Authors: Canzhi Chen, Zan Wang, Siqi Zhu, Qi Wu, Yixuan Li, Wei Liang

|Venue: arXiv preprint | Year: 2026

|arXiv: 2608.27371v1

Abstract

This paper introduces Embodied Scene Rearrangement Planning (ESRP), a novel task requiring embodied agents to rearrange furniture in 3D scenes to match a target configuration using only egocentric observations and a top-down target layout. Unlike prior rearrangement tasks, ESRP precludes global state access and introduces mutual object occlusions, reflecting the practical constraints of real-world robotic deployment. These factors make aligning partial egocentric observations with the global target layout particularly challenging for long-horizon planning. To facilitate research, we present ESRP-Bench, a comprehensive benchmark built on OmniGibson featuring over 5,400 scene pairs and 8,200 objects. We define three multi-level metrics to evaluate rearrangement quality and provide four baselines: a hierarchical task-and-motion planning method, a vision-language-model-based method, and two learning-based approaches (IL and RL). Experimental results demonstrate that current methods struggle to complete the task efficiently, highlighting ESRP as a challenging frontier for embodied agents in scene understanding and long-horizon task planning. This work serves as a stepping stone toward deploying intelligent agents in real-world scenarios. Project page: https://pie-lab.cn/ESRP/.

Key Contributions

• This paper introduces Embodied Scene Rearrangement Planning (ESRP), a novel task requiring embodi…
• Unlike prior rearrangement tasks, ESRP precludes global state access and introduces mutual object…
• These factors make aligning partial egocentric observations with the global target layout particu…
• To facilitate research, we present ESRP-Bench, a comprehensive benchmark built on OmniGibson feat…
• We define three multi-level metrics to evaluate rearrangement quality and provide four baselines:…

Topics

• vision
• reinforcement-learning
• planning
• benchmark

Code & Data

• GitHub repository: https://github.com/QwenLM/Qwen3-VL

BibTeX

@article{Chen2026_260827371v1,
  title     = {Embodied Scene Rearrangement Planning},
  author    = {Canzhi Chen and Zan Wang and Siqi Zhu and Qi Wu and Yixuan Li and Wei Liang},
  year      = {2026},
  eprint    = {2608.27371v1},
  archivePrefix = {arXiv},
  primaryClass  = {cs.RO},
  url       = {https://arxiv.org/abs/2608.27371v1}
}