# TRACKGRAPH: Online Open-Vocabulary 3D Scene Graphs via Image-Space Tracking

arXiv: https://arxiv.org/abs/2609.31005 · HTML: https://mlsanity.com/paper/2609.31005/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 26 references that are in the map.
- First posted on arXiv 25 Sep 2026.

## Builds on (most cited first, 25 of 25)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs](https://mlsanity.com/paper/2111.02114/) (2021)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [The Replica Dataset: A Digital Replica of Indoor Spaces](https://mlsanity.com/paper/1906.05797/) (2019)
- [ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenes](https://mlsanity.com/paper/2308.11417/) (2023)
- [OpenScene: 3D Scene Understanding with Open Vocabularies](https://mlsanity.com/paper/2211.15654/) (2022)
- [ConceptGraphs: Open-Vocabulary 3D Scene Graphs for Perception and Planning](https://mlsanity.com/paper/2309.16650/) (2023)
- [BoT-SORT: Robust Associations Multi-Pedestrian Tracking](https://mlsanity.com/paper/2206.14651/) (2022)
- [ConceptFusion: Open-set Multimodal 3D Mapping](https://mlsanity.com/paper/2302.07241/) (2023)
- [Fast Segment Anything](https://mlsanity.com/paper/2306.12156/) (2023)
- [Hierarchical Open-Vocabulary 3D Scene Graphs for Language-Grounded Robot Navigation](https://mlsanity.com/paper/2403.17846/) (2024)
- [Hydra: A Real-time Spatial Perception System for 3D Scene Graph Construction and Optimization](https://mlsanity.com/paper/2201.13360/) (2022)
- [Habitat-Matterport 3D Semantics Dataset](https://mlsanity.com/paper/2210.05633/) (2022)
- [Open-Fusion: Real-time Open-Vocabulary 3D Mapping and Queryable Scene Representation](https://mlsanity.com/paper/2310.03923/) (2023)
- [Open-Vocabulary Online Semantic Mapping for SLAM](https://mlsanity.com/paper/2411.15043/) (2024)
- [The Bare Necessities: Designing Simple, Effective Open-Vocabulary Scene Graphs](https://mlsanity.com/paper/2412.01539/) (2024)
- [OpenVox: Real-time Instance-level Open-vocabulary Probabilistic Voxel Representation](https://mlsanity.com/paper/2502.16528/) (2025)
- [FindAnything: Open-Vocabulary and Object-Centric Mapping for Robot Exploration in Any Environment](https://mlsanity.com/paper/2504.08603/) (2025)
- [OVI-MAP:Open-Vocabulary Instance-Semantic Mapping](https://mlsanity.com/paper/2603.26541/) (2026)
- [OpenTrack3D: Towards Accurate and Generalizable Open-Vocabulary 3D Instance Segmentation](https://mlsanity.com/paper/2512.03532/) (2025)
- [Think While You Map: Asynchronous Vision-Language Agents for Incremental 3D Scene Graphs](https://mlsanity.com/paper/2606.31471/) (2026)
- [OmniPlanner: Universal Exploration and Inspection Path Planning Across Robot Morphologies](https://mlsanity.com/paper/2603.04284/) (2026)
- [The Unified Autonomy Stack: Toward a Blueprint for Generalizable Robot Autonomy](https://mlsanity.com/paper/2605.12735/) (2026)

## Abstract (arXiv)

Open-vocabulary 3D maps enable robots to reason about previously unknown environments using natural language. However, existing systems typically segment every incoming image, associate detections with persistent 3D segments, and frequently perform costly Vision-Language (VL) inference. We present TRACKGRAPH, an online open-vocabulary system that maintains short-term 2D mask identity directly in the image stream before fusing segments into 3D. FastSAM masks and CLIP features are computed at sparse keyframes, while dense DINOv3 features are used to propagate masks at a high rate in between. The resulting tracked masks are fused into a class-agnostic 3D segment layer within a hierarchical scene graph, with 3D association handling tracking interruptions and long-term revisits. Compact multi-view CLIP embeddings enable open-vocabulary retrieval. Across Replica, ScanNet++, and HM3D, TRACKGRAPH achieves competitive open-vocabulary segmentation and retrieval against state-of-the-art mapping methods, including the highest synonym frequency on Replica (0.50). On the same NVIDIA A100, it is 1.7x faster and uses 3.3x less GPU memory than ViT-H OVI-MAP. Real-world quadruped deployments demonstrate onboard scene graph construction and object search at 7.5Hz, while recorded drone data is used to test the method under aerial viewpoints.