Deep Auxiliary Learning for Visual Localization and Odometry

TL;DR

This paper introduces VLocNet, a multitask CNN architecture for visual localization and odometry that outperforms existing deep learning methods and rivals traditional feature-based techniques.

Contribution

The paper presents a novel multitask CNN with a new loss function leveraging auxiliary learning for improved pose estimation accuracy.

Findings

VLocNet exceeds state-of-the-art deep architectures in global localization.

The model achieves competitive visual odometry performance.

Multitask learning with the Geometric Consistency Loss enhances localization accuracy.

Abstract

Localization is an indispensable component of a robot's autonomy stack that enables it to determine where it is in the environment, essentially making it a precursor for any action execution or planning. Although convolutional neural networks have shown promising results for visual localization, they are still grossly outperformed by state-of-the-art local feature-based techniques. In this work, we propose VLocNet, a new convolutional neural network architecture for 6-DoF global pose regression and odometry estimation from consecutive monocular images. Our multitask model incorporates hard parameter sharing, thus being compact and enabling real-time inference, in addition to being end-to-end trainable. We propose a novel loss function that utilizes auxiliary learning to leverage relative pose information during training, thereby constraining the search space to obtain consistent pose estimates. We evaluate our proposed VLocNet on indoor as well as outdoor datasets and show that even our single task model exceeds the performance of state-of-the-art deep architectures for global localization, while achieving competitive performance for visual odometry estimation. Furthermore, we present extensive experimental evaluations utilizing our proposed Geometric Consistency Loss that show the effectiveness of multitask learning and demonstrate that our model is the first deep learning technique to be on par with, and in some cases outperforms state-of-the-art SIFT-based approaches.