Recursive KalmanNet: Deep Learning-Augmented Kalman Filtering for State Estimation with Consistent Uncertainty Quantification

Hassan Mortada, Cyril Falcon, Yanis Kahil, Mathéo Clavaud, Jean-Philippe Michel
6/13/2025
eess.SPcs.LGstat.ML

No official code found. Best implementation: ixblue/RecursiveKalmanNet (24 stars, MIT, last commit 2025-09-11).

Abstract

State estimation in stochastic dynamical systems with noisy measurements is a challenge. While the Kalman filter is optimal for linear systems with independent Gaussian white noise, real-world conditions often deviate from these assumptions, prompting the rise of data-driven filtering techniques. This paper introduces Recursive KalmanNet, a Kalman-filter-informed recurrent neural network designed for accurate state estimation with consistent error covariance quantification. Our approach propagates error covariance using the recursive Joseph's formula and optimizes the Gaussian negative log-likelihood. Experiments with non-Gaussian measurement white noise demonstrate that our model outperforms both the conventional Kalman filter and an existing state-of-the-art deep learning based estimator.

View on arXivView PDF

Code Implementations(1)

ixblue/RecursiveKalmanNetCommunity100%
MIT

Deep Learning-Augmented Kalman Filtering for State Estimation with Consistent Uncertainty Quantification

241Python, Jupyter NotebookJun 13, 20251 years agoMIT

Similar Papers

Recursive KalmanNet: Analyse des capacités de généralisation d'un réseau de neurones récurrent guidé par un filtre de Kalman

85% match1 repo242025

Performance Evaluation of Deep Learning-Based State Estimation: A Comparative Study of KalmanNet

85% match2024

Uncertainty Representations in State-Space Layers for Deep Reinforcement Learning under Partial Observability

84% match2024

KalmanNet: Neural Network Aided Kalman Filtering for Partially Known Dynamics

81% match1 repo4602021

A Novel Neural Filter to Improve Accuracy of Neural Network Models of Dynamic Systems

80% match2024

GKNet: Graph Kalman Filtering and Model Inference via Model-based Deep Learning

80% match2025

Cite this paper

@article{mortada2025recursive,
  title  = {Recursive KalmanNet: Deep Learning-Augmented Kalman Filtering for State Estimation with Consistent Uncertainty Quantification},
  author = {Hassan Mortada and Cyril Falcon and Yanis Kahil and Mathéo Clavaud and Jean-Philippe Michel},
  year   = {2025},
  eprint = {2506.11639},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2506.11639v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2506.11639.svg)](https://codeofpaper.com/paper/2506.11639)

Discussion