# CoCo-InEKF: State Estimation with Learned Contact Covariances in Dynamic, Contact-Rich Scenarios

arXiv: https://arxiv.org/abs/2605.15122 · HTML: https://mlsanity.com/paper/2605.15122/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 43 references that are in the map.
- First posted on arXiv 14 May 2026.

## Builds on (most cited first, 24 of 24)

- [KalmanNet: Neural Network Aided Kalman Filtering for Partially Known Dynamics](https://mlsanity.com/paper/2107.10043/) (2021)
- [Concurrent Training of a Control Policy and a State Estimator for Dynamic and Robust Legged Locomotion](https://mlsanity.com/paper/2202.05481/) (2022)
- [RoNIN: Robust Neural Inertial Navigation in the Wild: Benchmark, Evaluations, and New Methods](https://mlsanity.com/paper/1905.12853/) (2019)
- [IONet: Learning to Cure the Curse of Drift in Inertial Odometry](https://mlsanity.com/paper/1802.02209/) (2018)
- [TLIO: Tight Learned Inertial Odometry](https://mlsanity.com/paper/2007.01867/) (2020)
- [Backprop KF: Learning Discriminative Deterministic State Estimators](https://mlsanity.com/paper/1605.07148/) (2016)
- [Multimodal Sensor Fusion with Differentiable Filters](https://mlsanity.com/paper/2010.13021/) (2020)
- [AI-IMU Dead-Reckoning](https://mlsanity.com/paper/1904.06064/) (2019)
- [Denoising IMU Gyroscopes with Deep Learning for Open-Loop Attitude Estimation](https://mlsanity.com/paper/2002.10718/) (2020)
- [Contact-Aided Invariant Extended Kalman Filtering for Robot State Estimation](https://mlsanity.com/paper/1904.09251/) (2019)
- [Learning Inertial Odometry for Dynamic Legged Robot State Estimation](https://mlsanity.com/paper/2111.00789/) (2021)
- [Legged Robot State Estimation using Invariant Kalman Filtering and Learned Contact Events](https://mlsanity.com/paper/2106.15713/) (2021)
- [Learned Inertial Odometry for Autonomous Drone Racing](https://mlsanity.com/paper/2210.15287/) (2022)
- [IMU Data Processing For Inertial Aided Navigation: A Recurrent Neural Network Based Approach](https://mlsanity.com/paper/2103.14286/) (2021)
- [Legged Robot State-Estimation Through Combined Forward Kinematic and Preintegrated Contact Factors](https://mlsanity.com/paper/1712.05873/) (2017)
- [Legged Robot State Estimation With Invariant Extended Kalman Filter Using Neural Measurement Network](https://mlsanity.com/paper/2402.00366/) (2024)
- [Robot Crash Course: Learning Soft and Stylized Falling](https://mlsanity.com/paper/2511.10635/) (2025)
- [Proprioceptive Invariant Robot State Estimation](https://mlsanity.com/paper/2311.04320/) (2023)
- [Legged Robot State Estimation Using Invariant Neural-Augmented Kalman Filter with a Neural Compensator](https://mlsanity.com/paper/2503.00344/) (2025)
- [State Estimation Transformers for Agile Legged Locomotion](https://mlsanity.com/paper/2410.13496/) (2024)
- [Robust State Estimation for Legged Robots with Dual Beta Kalman Filter](https://mlsanity.com/paper/2411.11483/) (2024)
- [Debiasing 6-DOF IMU via Hierarchical Learning of Continuous Bias Dynamics](https://mlsanity.com/paper/2504.09495/) (2025)
- [Adaptive Invariant Extended Kalman Filter for Legged Robot State Estimation](https://mlsanity.com/paper/2510.16755/) (2025)
- [InEKFormer: A Hybrid State Estimator for Humanoid Robots](https://mlsanity.com/paper/2511.16306/) (2025)

## Built on by (most cited first, 1 of 1)

- [Constrained Whole-Body Tracking for Humanoid Robots](https://mlsanity.com/paper/2606.00374/) (2026)

## Abstract (arXiv)

Robust state estimation for highly dynamic motion of legged robots remains challenging, especially in dynamic, contact-rich scenarios. Traditional approaches often rely on binary contact states that fail to capture the nuances of partial contact or directional slippage. This paper presents CoCo-InEKF, a differentiable invariant extended Kalman filter that utilizes continuous contact velocity covariances instead of binary contact states. These learned covariances allow the method to dynamically modulate contact confidence, accounting for more nuanced conditions ranging from firm contact to directional slippage or no contact. To predict these covariances for a set of predefined contact candidate points, we employ a lightweight neural network trained end-to-end using a state-error loss. This approach eliminates the need for heuristic ground-truth contact labels. In addition, we propose an automated contact candidate selection procedure and demonstrate that our method is insensitive to their exact placement. Experiments on a bipedal robot demonstrate a superior accuracy-efficiency tradeoff for linear velocity estimation, as well as improved filter consistency compared to baseline methods. This enables the robust execution of challenging motions, including dancing and complex ground interactions -- both in simulation and in the real world.