# Revisiting Articulated Parts Perception in Robot Manipulation

Source: [Revisiting Articulated Parts Perception in Robot Manipulation](<https://www.sota2.com/research/paper/6a279e3f90966ade919bf86c>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-06-15T06:00:01.130Z

Authors: Xiaoqian Wu, Yejie Guo, Xiaoyang Chen, Lixin Yang, Cewu Lu, Yong-Lu Li

Published: 2026-06-06T11:07:29.000Z

## Abstract

We are surrounded by various objects with movable, articulated parts, e.g., box, handle, door. An accurate and generalizable perception of articulated parts is essential to enhance robotic manipulation capabilities. Building on this need, recent efforts in articulated parts perception have followed two main directions: One line of work uses pose-based representation, which requires high manual cost; in parallel, affordance-based methods extract future object motion from point tracking without additional manual efforts, but suffer from low-quality data. In this paper, we propose a new representation of articulated parts, Geometric Primary Structure (GPS), an abstraction of the part geometry structure to balance scalability and quality. For efficient and scalable data collection, GPS is integrated with a portable Virtual Reality (VR) device and requires only one minute to annotate one object sequence. This direct human annotation provides higher quality than the estimated affordance. With this efficient VR-GPS system, we collect 41K frames for 234 objects across six part classes, and train a generalizable GPS model with a single RGB-D object image as input. For object manipulation, we deploy a heuristic policy based on GPS prediction. Without any in-domain fine-tuning, our method achieves an 73% success rate, covering 270 initial states for 9 objects. Our code, data and reusable tool are available at https://enlighten0707.github.io/gps.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Geometric Primitive Signature Prediction on RGBD-Art (test)](<https://www.sota2.com/research/sota/geometric-primitive-signature-prediction-on-rgbd-art-test.md>): Geometric Primitive Signature Prediction; RGBD-Art (test); AADE (deg): 19.29.

- [GPS Prediction on HOI4D (test)](<https://www.sota2.com/research/sota/gps-prediction-on-hoi4d-test.md>): GPS Prediction; HOI4D (test); AADE (Angular Error): 14.32.

- [Robot Manipulation on VR-GPS 9 objects (test)](<https://www.sota2.com/research/sota/robot-manipulation-on-vr-gps-9-objects-test.md>): Robot Manipulation; VR-GPS 9 objects (test); Box Success Rate: 93.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2606.08103>)

[PDF](<https://arxiv.org/pdf/2606.08103.pdf>)

[Implementation](<https://github.com/enlighten0707/Geometric_Primary_Structure>)
