Visual Language Maps for Robot Navigation

TLDR

Pretrained visual‑language models can align images with natural language, yet they lack the spatial precision of conventional geometric maps. The authors introduce VLMaps, a spatial map representation that fuses pretrained visual‑language features with a 3D reconstruction of the environment. VLMaps are autonomously constructed from robot video, allow natural‑language indexing without extra labels, and, when paired with large language models, translate commands into spatial navigation goals and support cross‑robot obstacle map sharing. Experiments in simulation and real‑world settings demonstrate that VLMaps enable navigation guided by more complex language instructions than prior methods. Videos are available at https://vlmaps.github.io.

Abstract

Grounding language to the visual observations of a navigating agent can be performed using off-the-shelf visual-language models pretrained on Internet-scale data (e.g., image captions). While this is useful for matching images to natural language descriptions of object goals, it remains disjoint from the process of mapping the environment, so that it lacks the spatial precision of classic geometric maps. To address this problem, we propose VLMaps, a spatial map representation that directly fuses pretrained visual-language features with a 3D reconstruction of the physical world. VLMaps can be autonomously built from video feed on robots using standard exploration approaches and enables natural language indexing of the map without additional labeled data. Specifically, when combined with large language models (LLMs), VLMaps can be used to (i) translate natural language commands into a sequence of open-vocabulary navigation goals (which, beyond prior work, can be spatial by construction, e.g., "in between the sofa and the TV" or "three meters to the right of the chair") directly localized in the map, and (ii) can be shared among multiple robots with different embodiments to generate new obstacle maps on-the-fly (by using a list of obstacle categories). Extensive experiments carried out in simulated and real-world environments show that VLMaps enable navigation according to more complex language instructions than existing methods. Videos are available at https://vlmaps.github.io.