SCOUT: Planning under Occlusion via Object-Centric World Model Rollouts
SCOUT: Planning under Occlusion via Object-Centric World Model Rollouts
This research presents SCOUT, an occlusion-aware planning framework that bridges high-level deliberation with grounded physical rollouts. By leveraging an object-centric scene representation to reconstruct hidden geometry and using a learned action-conditioned world model evaluated via Monte Carlo Tree Search, SCOUT outperforms Vision-Language-Action policies on 'Reveal' and 'Placement' tasks in cluttered environments. The framework was validated through real-world robotic manipulation in a cluttered shelf setting.
The paper introduces SCOUT, a planning framework that combines object-centric scene representation with Monte Carlo Tree Search and an action-conditioned world model to handle occlusions in cluttered environments, demonstrating improved performance over Vision-Language-Action policies on 'Reveal' and 'Placement' tasks.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Dataset details reported by the source
- Tasks
- Reveal Open source S4Placement Open source S4
- Embodiments
- robotic manipulation Open source S4
- paper: not_reported
What remains unknown
Dataset licensing and commercial-use rights are not established here.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
-conditioned world model. Evaluations on 'Reveal' and 'Placement' tasks show that SCOUT outperforms Vision-Language-Action (VLA) policies, embodied agents, and 3D world models, with significant gains in long-horizon tasks requiring reasoning over occluded geometry. We further validate the framework's robustness through real-world robotic manipulation in a cluttered shelf environment.
Open source S4
arXiv:2611.01600v1 Announce Type: new Abstract: Robots manipulating in cluttered spaces must often plan from partial views where foreground objects occlude the target, surrounding geometry, and usable free space. Reactive policies often miss these hidden constraints, while exhaustive search becomes computationally prohibitive in dense clutter. We present SCOUT, an occlusion-aware planning framework that bridges high-
Open source S2
Implications for data suppliers
RoboSignal interpretation and collection questions, not statements of buyer demand.
- Confirm the required data type and collection setting with the buyer; this source does not establish a complete collection specification.
- Check the exact dataset version, license, modalities and task coverage against the intended use.
- Validate demand and acceptance criteria with a buyer before scaling. Publication, popularity and a research result do not establish a purchase commitment.
Source：arXiv Robotics — research abstracts · arxiv.org