Kalman Filter
The Kalman filter is a powerful algorithm for estimating the state of a dynamic system from noisy measurements, aiming to minimize estimation error. Current research focuses on extending its application to nonlinear and non-Gaussian systems, often integrating it with deep learning models (like neural networks and transformers) or employing variants such as the extended Kalman filter (EKF), unscented Kalman filter (UKF), and invariant Kalman filter (InEKF) to handle complexities in diverse applications. This robust estimation technique finds widespread use in robotics, autonomous navigation, signal processing, and machine learning, improving accuracy and efficiency in various fields.
Papers
RINO: Accurate, Robust Radar-Inertial Odometry with Non-Iterative Estimation
Shuocheng Yang, Yueming Cao, Shengbo Li, Jianqiang Wang, Shaobing Xu
SP-VIO: Robust and Efficient Filter-Based Visual Inertial Odometry with State Transformation Model and Pose-Only Visual Description
Xueyu Du, Chengjun Ji, Lilian Zhang, Xinchan Luo, Huaiyi Zhang, Maosong Wang, Wenqi Wu, Jun Mao