UnDeepVO: Monocular Visual Odometry Through Unsupervised Deep Learning

Ruihao Li, Sen Wang, Zhiqiang Long, Dongbing Gu

2018 · 579 citations · 19 references

Concepts

TL;DR

UnDeepVO is a monocular visual odometry system that uniquely combines unsupervised deep learning with absolute scale recovery. The paper introduces UnDeepVO, a novel monocular visual odometry system. UnDeepVO estimates 6‑DoF pose and depth using deep neural networks trained on stereo pairs for scale recovery, with a loss function that exploits spatial and temporal dense information. On the KITTI dataset, UnDeepVO achieves competitive pose accuracy. Figure 1 shows the system overview.

Abstract

We propose a novel monocular visual odometry (VO) system called UnDeepVO in this paper. UnDeepVO is able to estimate the 6-DoF pose of a monocular camera and the depth of its view by using deep neural networks. There are two salient features of the proposed UnDeepVo:one is the unsupervised deep learning scheme, and the other is the absolute scale recovery. Specifically, we train UnDeepVoby using stereo image pairs to recover the scale but test it by using consecutive monocular images. Thus, UnDeepVO is a monocular system. The loss function defined for training the networks is based on spatial and temporal dense information. A system overview is shown in Fig. 1. The experiments on KITTI dataset show our UnDeepVO achieves good performance in terms of pose accuracy.

References

19