RAZER: Robust Accelerated Zero-Shot 3D Open-Vocabulary Panoptic Reconstruction with Spatio-Temporal Aggregation

--Mapping and understanding complex 3D environments is fundamental to how autonomous systems perceive and interact with the physical world, requiring both precise geometric reconstruction and rich semantic comprehension. While existing 3D semantic mapping systems excel at reconstructing and identifying predefined object instances, they lack the flexibility to efficiently build semantic maps with open-vocabulary during online operation. Although recent vision-language models have enabled open-vocabulary object recognition in 2D images, they haven't yet bridged the gap to 3D spatial understanding. The critical challenge lies in developing a training-free unified system that can simultaneously construct accurate 3D maps while maintaining semantic consistency and supporting natural language interactions in real time. In this paper, we develop a zero-shot framework that seamlessly integrates GPU-accelerated geometric reconstruction with open-vocabulary vision-language models through online instance-level semantic embedding fusion, guided by hierarchical object association with spatial indexing. Our training-free system achieves superior performance through incremental processing and unified geometric-semantic updates, while robustly handling 2D segmentation inconsistencies. The proposed general-purpose 3D scene understanding framework can be used for various tasks including zero-shot 3D instance retrieval, segmentation, and object detection to reason about previously unseen objects and interpret natural language queries. The project page is available at razer-3d.github.io. HE ability to create semantically meaningful 3D maps of dynamic environments is crucial for applications ranging from robotic navigation and manipulation to augmented reality and scene understanding. While significant advances have been made in geometric 3D reconstruction and 2D semantic understanding separately, combining these capabilities into a real-time system that can handle arbitrary objects and support natural language interactions remains a fundamental challenge in computer vision and robotics.