PointPilot: Unified Robotic Manipulation via Object-Centric 3D Point Flow Prediction
PointPilot: Unified Robotic Manipulation via Object-Centric 3D Point Flow Prediction
This research presents PointPilot, a task-focused manipulation framework that leverages 3D space point-flow prediction to handle domain variations across diverse embodiments, camera poses, and scene layouts.
The paper introduces PointPilot, a framework that achieves zero-shot generalization across different robot embodiments by predicting object-centric 3D point flows in a unified coordinate space, eliminating the need for separate action heads or fine-tuning.
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.
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ng the global scene, PointPilot only represents task-focused objects (robot grippers treated as objects), avoiding the prediction of potentially redundant scene parts unrelated to the manipulation tasks. Moreover, PointPilot aligns 3D spatial representation in one unified coordinate and supports closed-loop and fully automatic action execution across different manipulation modes from object point flow without a separ
Open source S3
trations from different embodiments. Extensive simulative and real world experiments validate our method. Notably, we demonstrate zero-shot embodiment generalization with PointPilot. A single trained policy can be deployed on previously unseen embodiments and successfully complete manipulation tasks without any additional fine-tuning.
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org