LS-VO: Learning Dense Optical Subspace for Robust Visual Odometry Estimation

Abstract

This work proposes a novel deep network architecture to solve the camera Ego-Motion estimation problem. A motion estimation network generally learns features similar to Optical Flow (OF) fields starting from sequences of images. This OF can be described by a lower dimensional latent space. Previous research has shown how to find linear approximations of this space. We propose to use an Auto-Encoder network to find a non-linear representation of the OF manifold. In addition, we propose to learn the latent space jointly with the estimation task, so that the learned OF features become a more robust description of the OF input. We call this novel architecture LS-VO. The experiments show that LS-VO achieves a considerable increase in performances in respect to baselines, while the number of parameters of the estimation network only slightly increases.

References in corpus (3)

• ORB-SLAM: a Versatile and Accurate Monocular SLAM System
• DeepVO: Towards End-to-End Visual Odometry with Deep Recurrent Convolutional Neural Networks
• DeMoN: Depth and Motion Network for Learning Monocular Stereo

Cited by in corpus (6)

• DeepIM: Deep Iterative Matching for 6D Pose Estimation
• Self-Supervised Scale Recovery for Monocular Depth and Egomotion Estimation
• ViPR: Visual-Odometry-aided Pose Regression for 6DoF Camera Localization
• Improving Variational Autoencoder based Out-of-Distribution Detection for Embedded Real-time Applications
• Monocular visual simultaneous localization and mapping: (r)evolution from geometry to deep learning-based pipelines
• Fusing Structure from Motion and Simulation-Augmented Pose Regression from Optical Flow for Challenging Indoor Environments