DeepUKF-VIN: Adaptively-tuned Deep Unscented Kalman Filter for 3D Visual-Inertial Navigation based on IMU-Vision-Net

TL;DR

This paper introduces DeepUKF-VIN, an adaptively-tuned deep unscented Kalman filter that fuses IMU and visual data for accurate 3D vehicle navigation in GPS-denied environments, demonstrating superior performance over standard methods.

Contribution

It proposes a novel deep learning-based adaptive tuning mechanism for the UKF in visual-inertial navigation, enhancing robustness and accuracy in 3D navigation tasks.

Findings

Demonstrates high estimation accuracy with real-world data

Shows superior performance compared to standard UKF

Ensures filter stability and rapid error attenuation

Abstract

This paper addresses the challenge of estimating the orientation, position, and velocity of a vehicle operating in three-dimensional (3D) space with six degrees of freedom (6-DoF). A Deep Learning-based Adaptation Mechanism (DLAM) is proposed to adaptively tune the noise covariance matrices of Kalman-type filters for the Visual-Inertial Navigation (VIN) problem, leveraging IMU-Vision-Net. Subsequently, an adaptively tuned Deep Learning Unscented Kalman Filter for 3D VIN (DeepUKF-VIN) is introduced to utilize the proposed DLAM, thereby robustly estimating key navigation components, including orientation, position, and linear velocity. The proposed DeepUKF-VIN integrates data from onboard sensors, specifically an inertial measurement unit (IMU) and visual feature points extracted from a camera, and is applicable for GPS-denied navigation. Its quaternion-based design effectively captures navigation nonlinearities and avoids the singularities commonly encountered with Euler-angle-based filters. Implemented in discrete space, the DeepUKF-VIN facilitates practical filter deployment. The filter's performance is evaluated using real-world data collected from an IMU and a stereo camera at low sampling rates. The results demonstrate filter stability and rapid attenuation of estimation errors, highlighting its high estimation accuracy. Furthermore, comparative testing against the standard Unscented Kalman Filter (UKF) in two scenarios consistently shows superior performance across all navigation components, thereby validating the efficacy and robustness of the proposed DeepUKF-VIN. Keywords: Deep Learning, Unscented Kalman Filter, Adaptive tuning, Estimation, Navigation, Unmanned Aerial Vehicle, Sensor-fusion.