Synthetic vs. Real Training Data for Visual Navigation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Suomela, Lauri, Arachchige, Sasanka Kuruppu, Torres, German F., Edelman, Harry, Kämäräinen, Joni-Kristian
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915815053328384
author Suomela, Lauri
Arachchige, Sasanka Kuruppu
Torres, German F.
Edelman, Harry
Kämäräinen, Joni-Kristian
author_facet Suomela, Lauri
Arachchige, Sasanka Kuruppu
Torres, German F.
Edelman, Harry
Kämäräinen, Joni-Kristian
contents This paper investigates how the performance of visual navigation policies trained in simulation compares to policies trained with real-world data. Performance degradation of simulator-trained policies is often significant when they are evaluated in the real world. However, despite this well-known sim-to-real gap, we demonstrate that simulator-trained policies can match the performance of their real-world-trained counterparts. Central to our approach is a navigation policy architecture that bridges the sim-to-real appearance gap by leveraging pretrained visual representations and runs real-time on robot hardware. Evaluations on a wheeled mobile robot show that the proposed policy, when trained in simulation, outperforms its real-world-trained version by 31 and the prior state-of-the-art methods by 50 points in navigation success rate. Policy generalization is verified by deploying the same model onboard a drone. Our results highlight the importance of diverse image encoder pretraining for sim-to-real generalization, and identify on-policy learning as a key advantage of simulated training over training with real data. Code, model checkpoints and multimedia materials are available at https://lasuomela.github.io/faint/
format Preprint
id arxiv_https___arxiv_org_abs_2509_11791
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Synthetic vs. Real Training Data for Visual Navigation
Suomela, Lauri
Arachchige, Sasanka Kuruppu
Torres, German F.
Edelman, Harry
Kämäräinen, Joni-Kristian
Robotics
Machine Learning
This paper investigates how the performance of visual navigation policies trained in simulation compares to policies trained with real-world data. Performance degradation of simulator-trained policies is often significant when they are evaluated in the real world. However, despite this well-known sim-to-real gap, we demonstrate that simulator-trained policies can match the performance of their real-world-trained counterparts. Central to our approach is a navigation policy architecture that bridges the sim-to-real appearance gap by leveraging pretrained visual representations and runs real-time on robot hardware. Evaluations on a wheeled mobile robot show that the proposed policy, when trained in simulation, outperforms its real-world-trained version by 31 and the prior state-of-the-art methods by 50 points in navigation success rate. Policy generalization is verified by deploying the same model onboard a drone. Our results highlight the importance of diverse image encoder pretraining for sim-to-real generalization, and identify on-policy learning as a key advantage of simulated training over training with real data. Code, model checkpoints and multimedia materials are available at https://lasuomela.github.io/faint/
title Synthetic vs. Real Training Data for Visual Navigation
topic Robotics
Machine Learning
url https://arxiv.org/abs/2509.11791