# SceneCut: Joint Geometric and Object Segmentation for Indoor Scenes

arXiv: https://arxiv.org/abs/1709.07158 · HTML: https://mlsanity.com/paper/1709.07158/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Sep 2017.

## Built on by (most cited first, 9 of 9)

- [Unseen Object Amodal Instance Segmentation via Hierarchical Occlusion Modeling](https://mlsanity.com/paper/2109.11103/) (2021)
- [Unsupervised Foreground Extraction via Deep Region Competition](https://mlsanity.com/paper/2110.15497/) (2021)
- [ICM-3D: Instantiated Category Modeling for 3D Instance Segmentation](https://mlsanity.com/paper/2108.11771/) (2021)
- [Unknown Object Segmentation from Stereo Images](https://mlsanity.com/paper/2103.06796/) (2021)
- [QueSTMaps: Queryable Semantic Topological Maps for 3D Scene Understanding](https://mlsanity.com/paper/2404.06442/) (2024)
- [RISeg: Robot Interactive Object Segmentation via Body Frame-Invariant Features](https://mlsanity.com/paper/2403.01731/) (2024)
- [Points2Vec: Unsupervised Object-level Feature Learning from Point Clouds](https://mlsanity.com/paper/2102.04136/) (2021)
- [rt-RISeg: Real-Time Model-Free Robot Interactive Segmentation for Active Instance-Level Object Understanding](https://mlsanity.com/paper/2507.10776/) (2025)
- [Fusing RGBD Tracking and Segmentation Tree Sampling for Multi-Hypothesis Volumetric Segmentation](https://mlsanity.com/paper/2104.00205/) (2021)

## Abstract (arXiv)

This paper presents SceneCut, a novel approach to jointly discover previously unseen objects and non-object surfaces using a single RGB-D image. SceneCut's joint reasoning over scene semantics and geometry allows a robot to detect and segment object instances in complex scenes where modern deep learning-based methods either fail to separate object instances, or fail to detect objects that were not seen during training. SceneCut automatically decomposes a scene into meaningful regions which either represent objects or scene surfaces. The decomposition is qualified by an unified energy function over objectness and geometric fitting. We show how this energy function can be optimized efficiently by utilizing hierarchical segmentation trees. Moreover, we leverage a pre-trained convolutional oriented boundary network to predict accurate boundaries from images, which are used to construct high-quality region hierarchies. We evaluate SceneCut on several different indoor environments, and the results show that SceneCut significantly outperforms all the existing methods.