Sequential Learning of Visual Tracking and Mapping Using Unsupervised Deep Neural Networks

TL;DR

This paper introduces an unsupervised deep learning-based SLAM system that integrates tracking, mapping, and optimization, demonstrating robustness and generality across diverse environments.

Contribution

It presents a novel end-to-end unsupervised learning framework for SLAM that includes uncertainty estimation and sequential training, enhancing robustness and adaptability.

Findings

Performs comparably with other learning-based VO methods

Effective in both indoor and outdoor environments

Uncertainty modeling improves robustness to noise

Abstract

We proposed an end-to-end deep learning-based simultaneous localization and mapping (SLAM) system following conventional visual odometry (VO) pipelines. The proposed method completes the SLAM framework by including tracking, mapping, and sequential optimization networks while training them in an unsupervised manner. Together with the camera pose and depth map, we estimated the observational uncertainty to make our system robust to noises such as dynamic objects. We evaluated our method using public indoor and outdoor datasets. The experiment demonstrated that our method works well in tracking and mapping tasks and performs comparably with other learning-based VO approaches. Notably, the proposed uncertainty modeling and sequential training yielded improved generality in a variety of environments.