Technology Category
Intelligent Robots: State EstimationMachine Learning: Calibration & Uncertainty QuantificationReasoning under Uncertainty: Relational Probabilistic Models
Recursive KalmanNet: Analyse des capacit'es de g'en'eralisation d'un r'eseau de neurones r'ecurrent guid'e par un filtre de Kalman
🤖 AI Summary
This work investigates the out-of-distribution (OOD) generalization of Recursive KalmanNet—specifically, its ability to robustly estimate both states and error covariance when test-time measurement dynamics significantly deviate from the training distribution. To address this, we propose a differentiable recurrent architecture that explicitly embeds the structural prior of Kalman filtering into a neural network, enabling joint end-to-end learning of system dynamics and noise statistics without requiring prior knowledge of noise distributions. By unifying sequential modeling with Bayesian filtering principles, the framework achieves data-driven, robust state estimation. Experiments demonstrate that our method maintains accurate state estimates and well-calibrated covariance predictions under unseen dynamics, substantially outperforming both classical Kalman filters (which assume known models and noise) and purely data-driven RNNs. It exhibits superior extrapolation capability and more reliable uncertainty quantification in OOD settings.
Technology Category
Application Category
📝 Abstract
The Recursive KalmanNet, recently introduced by the authors, is a recurrent neural network guided by a Kalman filter, capable of estimating the state variables and error covariance of stochastic dynamic systems from noisy measurements, without prior knowledge of the noise characteristics. This paper explores its generalization capabilities in out-of-distribution scenarios, where the temporal dynamics of the test measurements differ from those encountered during training. Le Recursive KalmanNet, r'ecemment introduit par les auteurs, est un r'eseau de neurones r'ecurrent guid'e par un filtre de Kalman, capable d'estimer les variables d''etat et la covariance des erreurs des syst`emes dynamiques stochastiques `a partir de mesures bruit'ees, sans connaissance pr'ealable des caract'eristiques des bruits. Cet article explore ses capacit'es de g'en'eralisation dans des sc'enarios hors distribution, o`u les dynamiques temporelles des mesures de test diff`erent de celles rencontr'ees `a l'entra^inement.
Problem
Research questions and friction points this paper is trying to address.
Estimating state variables without noise knowledge
Exploring generalization in out-of-distribution scenarios
Testing temporal dynamics differing from training
Innovation
Methods, ideas, or system contributions that make the work stand out.
Recurrent neural network guided by Kalman filter
Estimates state variables without noise knowledge
Explores generalization in out-of-distribution scenarios
🔎 Similar Papers
No similar papers found.
