OVI-MAP:Open-Vocabulary Instance-Semantic Mapping

2026-03-27

AI Summary

This work proposes a decoupled, incremental 3D instance-aware semantic mapping framework to address limitations in robustness, real-time performance, and open-vocabulary semantic reasoning of existing methods. Leveraging RGB-D inputs, the system constructs a class-agnostic 3D instance map online and extracts vision-language features (e.g., from CLIP) from automatically selected sparse keyframes to enable zero-shot semantic labeling and stable instance tracking. By decoupling geometric instance reconstruction from semantic inference, the approach avoids dense pixel-wise language fusion, significantly improving temporal consistency, scalability, and computational efficiency. Experiments demonstrate that the method outperforms state-of-the-art open-vocabulary mapping approaches on standard benchmarks while enabling real-time, high-precision semantic mapping.

Abstract

Incremental open-vocabulary 3D instance-semantic mapping is essential for autonomous agents operating in complex everyday environments. However, it remains challenging due to the need for robust instance segmentation, real-time processing, and flexible open-set reasoning. Existing methods often rely on the closed-set assumption or dense per-pixel language fusion, which limits scalability and temporal consistency. We introduce OVI-MAP that decouples instance reconstruction from semantic inference. We propose to build a class-agnostic 3D instance map that is incrementally constructed from RGB-D input, while semantic features are extracted only from a small set of automatically selected views using vision-language models. This design enables stable instance tracking and zero-shot semantic labeling throughout online exploration. Our system operates in real time and outperforms state-of-the-art open-vocabulary mapping baselines on standard benchmarks.

Problem

• open-vocabulary
• 3D instance-semantic mapping
• incremental mapping
• autonomous agents
• open-set reasoning

Innovation

• open-vocabulary mapping
• instance-semantic decoupling
• incremental 3D mapping
• vision-language models
• zero-shot semantic labeling