Learning Synthetic to Real Transfer for Localization and Navigational Tasks

TL;DR

This paper develops a simulation-to-real transfer pipeline for autonomous navigation, focusing on understanding the sim2real gap and designing modules for localization, planning, and local navigation using deep learning.

Contribution

It introduces a modular navigation pipeline trained in simulation with tailored datasets, emphasizing transferability and robustness for real-world autonomous navigation.

Findings

The pipeline effectively transfers from simulation to real-world environments.

A topological space representation improves generalization to new environments.

Localization as an image retrieval task enhances robustness and accuracy.

Abstract

Autonomous navigation consists in an agent being able to navigate without human intervention or supervision, it affects both high level planning and low level control. Navigation is at the crossroad of multiple disciplines, it combines notions of computer vision, robotics and control. This work aimed at creating, in a simulation, a navigation pipeline whose transfer to the real world could be done with as few efforts as possible. Given the limited time and the wide range of problematic to be tackled, absolute navigation performances while important was not the main objective. The emphasis was rather put on studying the sim2real gap which is one the major bottlenecks of modern robotics and autonomous navigation. To design the navigation pipeline four main challenges arise; environment, localization, navigation and planning. The iGibson simulator is picked for its photo-realistic textures and physics engine. A topological approach to tackle space representation was picked over metric approaches because they generalize better to new environments and are less sensitive to change of conditions. The navigation pipeline is decomposed as a localization module, a planning module and a local navigation module. These modules utilize three different networks, an image representation extractor, a passage detector and a local policy. The laters are trained on specifically tailored tasks with some associated datasets created for those specific tasks. Localization is the ability for the agent to localize itself against a specific space representation. It must be reliable, repeatable and robust to a wide variety of transformations. Localization is tackled as an image retrieval task using a deep neural network trained on an auxiliary task as a feature descriptor extractor. The local policy is trained with behavioral cloning from expert trajectories gathered with ROS navigation stack.