Reinforcement Learning Meets Visual Odometry

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Messikommer, Nico, Cioffi, Giovanni, Gehrig, Mathias, Scaramuzza, Davide
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916331863932928
author Messikommer, Nico
Cioffi, Giovanni
Gehrig, Mathias
Scaramuzza, Davide
author_facet Messikommer, Nico
Cioffi, Giovanni
Gehrig, Mathias
Scaramuzza, Davide
contents Visual Odometry (VO) is essential to downstream mobile robotics and augmented/virtual reality tasks. Despite recent advances, existing VO methods still rely on heuristic design choices that require several weeks of hyperparameter tuning by human experts, hindering generalizability and robustness. We address these challenges by reframing VO as a sequential decision-making task and applying Reinforcement Learning (RL) to adapt the VO process dynamically. Our approach introduces a neural network, operating as an agent within the VO pipeline, to make decisions such as keyframe and grid-size selection based on real-time conditions. Our method minimizes reliance on heuristic choices using a reward function based on pose error, runtime, and other metrics to guide the system. Our RL framework treats the VO system and the image sequence as an environment, with the agent receiving observations from keypoints, map statistics, and prior poses. Experimental results using classical VO methods and public benchmarks demonstrate improvements in accuracy and robustness, validating the generalizability of our RL-enhanced VO approach to different scenarios. We believe this paradigm shift advances VO technology by eliminating the need for time-intensive parameter tuning of heuristics.
format Preprint
id arxiv_https___arxiv_org_abs_2407_15626
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Reinforcement Learning Meets Visual Odometry
Messikommer, Nico
Cioffi, Giovanni
Gehrig, Mathias
Scaramuzza, Davide
Computer Vision and Pattern Recognition
Robotics
Visual Odometry (VO) is essential to downstream mobile robotics and augmented/virtual reality tasks. Despite recent advances, existing VO methods still rely on heuristic design choices that require several weeks of hyperparameter tuning by human experts, hindering generalizability and robustness. We address these challenges by reframing VO as a sequential decision-making task and applying Reinforcement Learning (RL) to adapt the VO process dynamically. Our approach introduces a neural network, operating as an agent within the VO pipeline, to make decisions such as keyframe and grid-size selection based on real-time conditions. Our method minimizes reliance on heuristic choices using a reward function based on pose error, runtime, and other metrics to guide the system. Our RL framework treats the VO system and the image sequence as an environment, with the agent receiving observations from keypoints, map statistics, and prior poses. Experimental results using classical VO methods and public benchmarks demonstrate improvements in accuracy and robustness, validating the generalizability of our RL-enhanced VO approach to different scenarios. We believe this paradigm shift advances VO technology by eliminating the need for time-intensive parameter tuning of heuristics.
title Reinforcement Learning Meets Visual Odometry
topic Computer Vision and Pattern Recognition
Robotics
url https://arxiv.org/abs/2407.15626