PIVO: Probabilistic Inertial-Visual Odometry for Occlusion-Robust Navigation

TL;DR

This paper introduces PIVO, a probabilistic visual-inertial odometry method that fuses smartphone IMU and camera data, improving robustness in occlusion and feature-scarce environments through comprehensive uncertainty propagation.

Contribution

It presents a novel information fusion framework with a sequential inference scheme that models all cross-terms, enhancing robustness against occlusion and poor feature conditions.

Findings

Demonstrates robustness in occlusion scenarios

Outperforms Tango device in accuracy

Effective on iPhone and EuRoC datasets

Abstract

This paper presents a novel method for visual-inertial odometry. The method is based on an information fusion framework employing low-cost IMU sensors and the monocular camera in a standard smartphone. We formulate a sequential inference scheme, where the IMU drives the dynamical model and the camera frames are used in coupling trailing sequences of augmented poses. The novelty in the model is in taking into account all the cross-terms in the updates, thus propagating the inter-connected uncertainties throughout the model. Stronger coupling between the inertial and visual data sources leads to robustness against occlusion and feature-poor environments. We demonstrate results on data collected with an iPhone and provide comparisons against the Tango device and using the EuRoC data set.