ReVNM: Learning-Based Visual Navigation from a Remote Camera

2026-09-23

AI Summary

This study addresses the challenges of obstacle avoidance and training data scarcity arising from the limited field of view in remote cameras. We propose a map-free visual navigation method that utilizes a single remote surveillance camera as both the observation source and an implicit environment representation. Training data are synthesized via a randomly generated world policy, and an exocentric-to-egocentric transformation module is designed to predict depth information. The proposed model achieves cross-domain generalization without requiring fine-tuning on real-world scenes. Both simulation and real-world experiments demonstrate that this approach effectively enables autonomous, robust robot navigation and collision avoidance from a remote perspective.

Abstract

Visual Navigation Models (VNMs) enable robots to navigate from egocentric visual observations without geometric localization and planning, but long-range navigation still requires pre-built maps. This paper presents the Remote Visual Navigation Model (ReVNM), which uses a single remote surveillance camera to serve as both an observation source and an implicit environmental map for visual navigation. While the use of remote cameras could eliminate the need for pre-built maps as well as onboard vision processing, their limited field of view instead of egocentric observations makes it hard to achieve collision-free navigation. The lack of existing data with diverse remote viewpoints, which are crucial for training robust VNMs, further complicates the challenge. In this work, we propose a learning-by-synthesis approach to address this two-fold challenge. Our ReVNM extends a state-of-the-art VNM architecture with an exocentric-to-egocentric (exo2ego) module that predicts an egocentric depth observation from remote-camera observations. This helps the VNM to plan a path while considering obstacles in front of the robot. Trained only on randomly generated worlds with diverse obstacle layouts and camera viewpoints, ReVNM can generalize well to real robot navigation without additional fine-tuning. Experiments in both simulation and real-world environments confirmed the effectiveness of the proposed approach.

Problem

• Visual Navigation
• Remote Camera
• Collision-free Navigation
• Exocentric Observation
• Pre-built Map

Innovation

• Remote Visual Navigation
• Exocentric-to-Egocentric
• Learning-by-Synthesis
• Depth Prediction
• Sim-to-Real Generalization