Scale-aware direct monocular odometry

TL;DR

This paper introduces a scale-aware monocular odometry framework that integrates multi-view depth prediction and robust optimization, significantly improving accuracy and eliminating scale drift in monocular SLAM.

Contribution

It proposes a novel depth residual formulation and a robust cost function, enabling effective multi-view depth integration without relying on specific neural networks.

Findings

Outperforms classic monocular SLAM by 5-9 times in accuracy.

Achieves accuracy comparable to stereo systems.

Validates the approach on the KITTI dataset with multiple neural networks.

Abstract

We present a generic framework for scale-aware direct monocular odometry based on depth prediction from a deep neural network. In contrast with previous methods where depth information is only partially exploited, we formulate a novel depth prediction residual which allows us to incorporate multi-view depth information. In addition, we propose to use a truncated robust cost function which prevents considering inconsistent depth estimations. The photometric and depth-prediction measurements are integrated into a tightly-coupled optimization leading to a scale-aware monocular system which does not accumulate scale drift. Our proposal does not particularize for a concrete neural network, being able to work along with the vast majority of the existing depth prediction solutions. We demonstrate the validity and generality of our proposal evaluating it on the KITTI odometry dataset, using two publicly available neural networks and comparing it with similar approaches and the state-of-the-art for monocular and stereo SLAM. Experiments show that our proposal largely outperforms classic monocular SLAM, being 5 to 9 times more precise, beating similar approaches and having an accuracy which is closer to that of stereo systems.