QueryGaussian: Scalable and Training-Free Open-Vocabulary 3D Instance Retrieval
Quick Answer
QueryGaussian introduces a training-free framework for scalable open-vocabulary 3D instance retrieval, achieving over 70% GPU memory reduction and 180x faster inference.
Quick Take
This method leverages pre-trained 2D models for semantic interpretation, enabling efficient retrieval in city-scale environments with millions of instances.
Key Points
- QueryGaussian reduces GPU memory usage by over 70% compared to existing methods.
- Achieves 180x faster inference times, making it suitable for real-time applications.
- Utilizes pre-trained 2D vision models for effective semantic understanding.
- Decouples semantic understanding from geometric representation for improved efficiency.
- Enables retrieval in city-scale scenes with tens of millions of instances.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Efficiently retrieving specific 3D instances from large-scale scenes via natural language prompts remains a formidable challenge in multimedia analysis. Existing approaches predominantly follow a "scene-level embedding" paradigm, which requires distilling high-dimensional semantic features into every 3D primitive. This strategy suffers from a fundamental architectural bottleneck: memory and computational costs scale linearly with scene complexity, inevitably triggering out-of-memory (OOM) failures in city-scale environments. To address this barrier, we propose QueryGaussian, a training-free framework for expeditious and scalable open-vocabulary 3D instance retrieval. Unlike holistic semantic distillation, QueryGaussian employs an instance-level query mechanism that decouples semantic understanding from geometric representation. Specifically, we leverage pre-trained 2D vision models to interpret user prompts and lift segmentation masks into 3D via a concurrent maximum-weight association strategy, ensuring semantic-visual consistency. To mitigate projection ambiguity, we introduce a temporal fusion module with multi-stage adaptive density clustering. Experimental results demonstrate that QueryGaussian not only matches the accuracy of state-of-the-art methods but also delivers a decisive efficiency leap, reducing GPU memory usage by over 70% and accelerating inference by 180x. Crucially, QueryGaussian enables expeditious instance retrieval on city-scale scenes containing tens of millions of Gaussians using consumer-grade hardware.
| Comments: | 8 pages, 4 figures, 6 tables. Accepted to the 2026 IEEE International Conference on Systems, Man, and Cybernetics (SMC 2026) |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2606.19733 [cs.CV] |
| (or arXiv:2606.19733v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2606.19733 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Xiuyuan Zhu [view email]
[v1]
Thu, 18 Jun 2026 02:57:35 UTC (5,230 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
ProMoE-FL introduces a Prototype-conditioned Mixture-of-Experts framework for multimodal federated learning, effectively addressing missing modalities. It outperforms existing methods on four chest X-ray datasets, demonstrating superior feature synthesis capabilities in both homogeneous and heterogeneous settings.