AdaVLN: Towards Visual Language Navigation in Continuous Indoor Environments with Moving Humans

TL;DR

AdaVLN extends visual language navigation to dynamic indoor environments with moving humans, introducing a new simulator, datasets, and mechanisms to handle real-world complexities and improve reproducibility.

Contribution

We propose AdaVLN, a novel task and environment for navigation amidst moving humans, along with datasets and a freeze-time mechanism to facilitate research and reproducibility.

Findings

Baseline models face challenges with dynamic obstacles.

AdaVLN bridges the sim-to-real gap in VLN.

New datasets and simulator support dynamic environment research.

Abstract

Visual Language Navigation is a task that challenges robots to navigate in realistic environments based on natural language instructions. While previous research has largely focused on static settings, real-world navigation must often contend with dynamic human obstacles. Hence, we propose an extension to the task, termed Adaptive Visual Language Navigation (AdaVLN), which seeks to narrow this gap. AdaVLN requires robots to navigate complex 3D indoor environments populated with dynamically moving human obstacles, adding a layer of complexity to navigation tasks that mimic the real-world. To support exploration of this task, we also present AdaVLN simulator and AdaR2R datasets. The AdaVLN simulator enables easy inclusion of fully animated human models directly into common datasets like Matterport3D. We also introduce a "freeze-time" mechanism for both the navigation task and simulator, which pauses world state updates during agent inference, enabling fair comparisons and experimental reproducibility across different hardware. We evaluate several baseline models on this task, analyze the unique challenges introduced by AdaVLN, and demonstrate its potential to bridge the sim-to-real gap in VLN research.