![[2602.14193] Learning Part-Aware Dense 3D Feature Field for Generalizable Articulated Object Manipulation](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.14193] Learning Part-Aware Dense 3D Feature Field for Generalizable Articulated Object Manipulation
Summary
The paper presents a novel Part-Aware 3D Feature Field (PA3FF) for enhancing robotic manipulation of articulated objects, addressing challenges in generalization across diverse shapes and categories.
Why It Matters
This research is significant as it tackles the critical issue of generalizing robotic manipulation across various articulated objects, which is essential for real-world applications in robotics. By focusing on functional parts, the proposed method enhances the efficiency and effectiveness of robotic systems, paving the way for more advanced applications in automation and AI.
Key Takeaways
- PA3FF improves generalization in robotic manipulation by focusing on functional parts.
- The method outperforms existing 2D and 3D representations in various tasks.
- Part-Aware Diffusion Policy (PADP) enhances sample efficiency in imitation learning.
- The approach supports diverse downstream applications, including segmentation and correspondence learning.
- Utilizing a large-scale labeled dataset, PA3FF is trained via contrastive learning.
Computer Science > Robotics arXiv:2602.14193 (cs) [Submitted on 15 Feb 2026] Title:Learning Part-Aware Dense 3D Feature Field for Generalizable Articulated Object Manipulation Authors:Yue Chen, Muqing Jiang, Kaifeng Zheng, Jiaqi Liang, Chenrui Tie, Haoran Lu, Ruihai Wu, Hao Dong View a PDF of the paper titled Learning Part-Aware Dense 3D Feature Field for Generalizable Articulated Object Manipulation, by Yue Chen and 7 other authors View PDF HTML (experimental) Abstract:Articulated object manipulation is essential for various real-world robotic tasks, yet generalizing across diverse objects remains a major challenge. A key to generalization lies in understanding functional parts (e.g., door handles and knobs), which indicate where and how to manipulate across diverse object categories and shapes. Previous works attempted to achieve generalization by introducing foundation features, while these features are mostly 2D-based and do not specifically consider functional parts. When lifting these 2D features to geometry-profound 3D space, challenges arise, such as long runtimes, multi-view inconsistencies, and low spatial resolution with insufficient geometric information. To address these issues, we propose Part-Aware 3D Feature Field (PA3FF), a novel dense 3D feature with part awareness for generalizable articulated object manipulation. PA3FF is trained by 3D part proposals from a large-scale labeled dataset, via a contrastive learning formulation. Given point clouds as input,...