Self-Supervised Visual Place Recognition Learning in Mobile Robots

TL;DR

This paper introduces a self-supervised learning method for visual place recognition in mobile robots, using GPS data to bootstrap labels and improve loop-closure detection accuracy.

Contribution

It presents a novel self-supervised approach that learns a specialized distance metric for visual place recognition using existing GPS-aided navigation.

Findings

Learned distance metric improves loop-closure identification.

Self-supervised approach outperforms traditional methods.

Effective with various CNN-based image descriptors.

Abstract

Place recognition is a critical component in robot navigation that enables it to re-establish previously visited locations, and simultaneously use this information to correct the drift incurred in its dead-reckoned estimate. In this work, we develop a self-supervised approach to place recognition in robots. The task of visual loop-closure identification is cast as a metric learning problem, where the labels for positive and negative examples of loop-closures can be bootstrapped using a GPS-aided navigation solution that the robot already uses. By leveraging the synchronization between sensors, we show that we are able to learn an appropriate distance metric for arbitrary real-valued image descriptors (including state-of-the-art CNN models), that is specifically geared for visual place recognition in mobile robots. Furthermore, we show that the newly learned embedding can be particularly powerful in disambiguating visual scenes for the task of vision-based loop-closure identification in mobile robots.