IMU

• End-to-end learning framework for real-time inertial attitude estimation. Generalized across various sampling rates. RNN-CNN networks employed to learn motion characteristics, noise, and bias. Proposed approach outperforms traditional algorithms and other deepOutperforms traditional algorithms in terms of accuracy up to 40 Evaluated using seven datasets, totaling 120 h and 200 kilometers of IMU measurements.

Introduction RepoIMU T-stick RepoIMU T-pendulum Sassari OxIOD MAV Dataset EuRoC MAV TUM-VI KITTI RIDI RoNIN BROAD References Introduction IMU Datasets are used to evaluate the performance of the attitude estimation algorithms.