# Occupancy-Grounded Room Segmentation for Hierarchical 3D Scene Graphs

Source: [Occupancy-Grounded Room Segmentation for Hierarchical 3D Scene Graphs](<https://www.sota2.com/research/paper/6a2f870dd33dfb23e783bf36>)

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

Data updated: 2026-06-16T05:00:05.515Z

Authors: Carlos Cueto Zumaya, Iacopo Catalano, Jorge Pe\\~na-Queralta, Wallace Moreira Bessa

Published: 2026-06-11T09:36:15.000Z

## Abstract

Hierarchical 3D scene graphs (3DSGs) for indoor robots organize geometric and semantic information across spatial scales, with a room layer that connects object-level perception to room-scale reasoning. Existing systems construct this layer from different spatial substrates (\\eg{} place clusters, wall planes, or segmentation outputs), and as a result, room nodes are not evaluated on a common geometric criterion. We present an occupancy-grounded 3DSG pipeline in which room nodes are anchored to tracked free-space regions derived from occupancy decomposition, giving each room an explicit polygonal footprint. We evaluate the pipeline on 12 Matterport3D scenes by matching predicted room polygons to annotated room instances and compare against Hydra, a representative state-of-the-art place-connectivity baseline. The results show that occupancy-grounded anchoring recovers substantially more room instances than place-connectivity construction, at the cost of lower precision, and that wall-accurate room boundaries remain an open problem for both methods. Code is available at https://github.com/crcz25/OccuSG.

## 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.

- [Room-instance recovery on Matterport3D (12 scenes)](<https://www.sota2.com/research/sota/room-instance-recovery-on-matterport3d-12-scenes.md>): Room-instance recovery; Matterport3D (MP3D) (12 scenes); Recall (R): 66.7.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2606.13727>)

[PDF](<https://arxiv.org/pdf/2606.13727.pdf>)
