# Amodal 3D Reconstruction for Robotic Manipulation via Stability and Connectivity

arXiv: https://arxiv.org/abs/2009.13146 · HTML: https://mlsanity.com/paper/2009.13146/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Sep 2020.

## Built on by (most cited first, 15 of 15)

- [Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenes](https://mlsanity.com/paper/2103.14127/) (2021)
- [FiG-NeRF: Figure-Ground Neural Radiance Fields for 3D Object Category Modelling](https://mlsanity.com/paper/2104.08418/) (2021)
- [Unseen Object Amodal Instance Segmentation via Hierarchical Occlusion Modeling](https://mlsanity.com/paper/2109.11103/) (2021)
- [DDGC: Generative Deep Dexterous Grasping in Clutter](https://mlsanity.com/paper/2103.04783/) (2021)
- [GSWorld: Closed-Loop Photo-Realistic Simulation Suite for Robotic Manipulation](https://mlsanity.com/paper/2510.20813/) (2025)
- [Teaching Robots to Build Simulations of Themselves](https://mlsanity.com/paper/2311.12151/) (2023)
- [PokeFlex: A Real-World Dataset of Volumetric Deformable Objects for Robotics](https://mlsanity.com/paper/2410.07688/) (2024)
- [Picasso: Holistic Scene Reconstruction with Physics-Constrained Sampling](https://mlsanity.com/paper/2602.08058/) (2026)
- [WALT3D: Generating Realistic Training Data from Time-Lapse Imagery for Reconstructing Dynamic Objects under Occlusion](https://mlsanity.com/paper/2403.19022/) (2024)
- [WorldSculpt: Generating Compositional Worlds from Grounded Videos](https://mlsanity.com/paper/2609.05416/) (2026)
- [Object Reconstruction under Occlusion with Generative Priors and Contact-induced Constraints](https://mlsanity.com/paper/2512.05079/) (2025)
- [Robust Bayesian Scene Reconstruction with Retrieval-Augmented Priors for Precise Grasping and Planning](https://mlsanity.com/paper/2411.19461/) (2024)
- [Exploiting Radiance Fields for Grasp Generation on Novel Synthetic Views](https://mlsanity.com/paper/2505.11467/) (2025)
- [BIMStruct3D: A Fully Automated Hybrid Learning Scan-to-BIM Pipeline with Integrated Topology Refinement](https://mlsanity.com/paper/2604.24311/) (2026)
- [CODA: Depth-Aligned Scene Completion and Object Decomposition from a Single RGB-D Image](https://mlsanity.com/paper/2609.25654/) (2026)

## Abstract (arXiv)

Learning-based 3D object reconstruction enables single- or few-shot estimation of 3D object models. For robotics, this holds the potential to allow model-based methods to rapidly adapt to novel objects and scenes. Existing 3D reconstruction techniques optimize for visual reconstruction fidelity, typically measured by chamfer distance or voxel IOU. We find that when applied to realistic, cluttered robotics environments, these systems produce reconstructions with low physical realism, resulting in poor task performance when used for model-based control. We propose ARM, an amodal 3D reconstruction system that introduces (1) a stability prior over object shapes, (2) a connectivity prior, and (3) a multi-channel input representation that allows for reasoning over relationships between groups of objects. By using these priors over the physical properties of objects, our system improves reconstruction quality not just by standard visual metrics, but also performance of model-based control on a variety of robotics manipulation tasks in challenging, cluttered environments. Code is available at github.com/wagnew3/ARM.