With the recent advance of deep learning based object recognition and estimation, it is possible to consider level SLAM where pose each estimated in process. In this letter, on a novel Lie group structure, right invariant extended Kalman filter (RI-EKF) for proposed. The observability analysis shows that proposed algorithm automatically maintains correct unobservable subspace, while standard EK...

In literature, Extended Object Tracking (EOT) algorithms developed for autonomous driving predominantly provide obstacles state estimation in cartesian coordinates the Vehicle Reference Frame. However, many scenarios, representation road-aligned curvilinear is preferred when implementing subsystems like cruise control, lane-keeping assist, platooning, etc. This paper proposes a Gaussian Mixture...

We present an alternative inference framework for the Gaussian process-based extended object tracking (GPEOT) models. The method provides approximate solution to Bayesian filtering problem in GPEOT by relying on a new measurement update, which we derive using variational Bayes techniques. resulting algorithm effectively computes posterior densities of kinematic and extent states. conduct variou...