Nine-Axis IMU-based Extended inertial odometry neural network

作者：

Highlights：

• Inertial odometry neural network that learns gravity acceleration is proposed.

• Pose-TuningNet for learning current pose information in network parts is proposed.

• It improves drift by estimating orientation via gravity acceleration and geomagnetism.

• Extended IONet has about 40% less trajectory estimation error than the existing model.

摘要

•Inertial odometry neural network that learns gravity acceleration is proposed.•Pose-TuningNet for learning current pose information in network parts is proposed.•It improves drift by estimating orientation via gravity acceleration and geomagnetism.•Extended IONet has about 40% less trajectory estimation error than the existing model.

论文关键词：Extended inertial odometry neural network,Inertial measurement unit,Drift,Pose-TuningNet

论文评审过程：Received 18 September 2020, Revised 18 March 2021, Accepted 17 April 2021, Available online 21 April 2021, Version of Record 29 April 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115075