VANDERER: Map-Free Exploration using Future-Aware and Visual-Curiosity-Guided Diffusion Policy

2026-06-12

AI Summary

This work addresses the challenge of robotic exploration under perception-limited conditions where only a monocular camera is available, rendering conventional map-based methods ineffective due to their inability to construct accurate occupancy maps. To overcome this limitation, the authors propose a novel exploration framework that eschews explicit mapping by integrating visual curiosity mechanisms with a pretrained diffusion policy. The approach leverages a navigation world model and a monocular visual action evaluation module to drive exploration through curiosity-guided action optimization. Evaluated across multiple simulated environments, the method explores on average 13.4% more area than NoMaD and significantly outperforms existing baselines. Furthermore, the study uncovers an intrinsic relationship between visual and geometric curiosity in outdoor scenes, offering new insights into curiosity-driven exploration.

Abstract

Mobile agents require efficient exploration strategies to map unseen environments and autonomously plan tasks. Traditional methods rely on generating occupancy maps and optimizing the sequence in which unexplored regions are visited. However, in sensor-constrained settings, such as those limited to monocular cameras, generating accurate occupancy maps is challenging. To address this, we propose VANDERER, an exploration framework that leverages a Visual Curiosity Module (VCM) to guide pre-trained diffusion policies using only monocular image data. This curiosity module predicts the outcomes of proposed actions via a navigation world model and evaluates them through a curiosity cost. The cost then guides the diffusion process toward generating actions that maximize exploration. Evaluated across diverse simulated environments, VANDERER consistently outperforms established baselines, exploring an average of 13.4% more area than NoMaD. Our results reveal a direct correlation between visual and geometric curiosity in outdoor environments, demonstrating that VANDERER can effectively leverage this relationship for efficient exploration using sensor-constrained agents.

Problem

• map-free exploration
• sensor-constrained agents
• monocular vision
• efficient exploration
• occupancy mapping

Innovation

• map-free exploration
• diffusion policy
• visual curiosity
• monocular navigation
• world model