# Dr-BA: Separable Optimization for Direct Radar Bundle Adjustment & Localization

arXiv: https://arxiv.org/abs/2605.07041 · HTML: https://mlsanity.com/paper/2605.07041/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 37 references that are in the map.
- First posted on arXiv 7 May 2026.

## Builds on (most cited first, 22 of 22)

- [Boreas: A Multi-Season Autonomous Driving Dataset](https://mlsanity.com/paper/2203.10168/) (2022)
- [Scan Context++: Structural Place Recognition Robust to Rotation and Lateral Variations in Urban Environments](https://mlsanity.com/paper/2109.13494/) (2021)
- [A New Wave in Robotics: Survey on Recent mmWave Radar Applications in Robotics](https://mlsanity.com/paper/2305.01135/) (2023)
- [Masking by Moving: Learning Distraction-Free Radar Odometry from Pose Information](https://mlsanity.com/paper/1909.03752/) (2019)
- [Are We Ready for Radar to Replace Lidar in All-Weather Mapping and Localization?](https://mlsanity.com/paper/2203.10174/) (2022)
- [An Efficient Global Optimality Certificate for Landmark-Based SLAM](https://mlsanity.com/paper/2206.12961/) (2022)
- [Lidar-level localization with radar? The CFEAR approach to accurate, fast and robust large-scale radar odometry in diverse environments](https://mlsanity.com/paper/2211.02445/) (2022)
- [RaPlace: Place Recognition for Imaging Radar using Radon Transform and Mutable Threshold](https://mlsanity.com/paper/2307.04321/) (2023)
- [A Normal Distribution Transform-Based Radar Odometry Designed For Scanning and Automotive Radars](https://mlsanity.com/paper/2103.07908/) (2021)
- [TBV Radar SLAM -- trust but verify loop candidates](https://mlsanity.com/paper/2301.04397/) (2023)
- [Radar Odometry for Autonomous Ground Vehicles: A Survey of Methods and Datasets](https://mlsanity.com/paper/2307.07861/) (2023)
- [Continuous-Time Radar-Inertial and Lidar-Inertial Odometry using a Gaussian Process Motion Prior](https://mlsanity.com/paper/2402.06174/) (2024)
- [Are Doppler Velocity Measurements Useful for Spinning Radar Odometry?](https://mlsanity.com/paper/2404.01537/) (2024)
- [Boreas Road Trip: A Multi-Sensor Autonomous Driving Dataset on Challenging Roads](https://mlsanity.com/paper/2602.16870/) (2026)
- [ORORA: Outlier-Robust Radar Odometry](https://mlsanity.com/paper/2303.01876/) (2023)
- [DSR: Direct Simultaneous Registration for Multiple 3D Images](https://mlsanity.com/paper/2105.10087/) (2021)
- [The Finer Points: A Systematic Comparison of Point-Cloud Extractors for Radar Odometry](https://mlsanity.com/paper/2409.12256/) (2024)
- [2Fast-2Lamaa: Large-Scale Lidar-Inertial Localization and Mapping with Continuous Distance Fields](https://mlsanity.com/paper/2410.05433/) (2024)
- [DRO: Doppler-Aware Direct Radar Odometry](https://mlsanity.com/paper/2504.20339/) (2025)
- [XPRESS: X-Band Radar Place Recognition via Elliptical Scan Shaping](https://mlsanity.com/paper/2511.08863/) (2025)
- [Sparse Variable Projection in Robotic Perception: Exploiting Separable Structure for Efficient Nonlinear Optimization](https://mlsanity.com/paper/2512.07969/) (2025)
- [Dr-PoGO: Direct Radar Pose-Graph Optimization](https://mlsanity.com/paper/2605.04806/) (2026)

## Abstract (arXiv)

This paper introduces Dr-BA, a first-of-its-kind radar bundle adjustment (BA) framework that operates directly on 2D spinning radar intensity images. Unlike camera or lidar sensors, radar is largely unaffected by precipitation, making it a critical modality for autonomous systems that require all-weather robustness. Existing state estimation approaches using spinning radar typically extract sparse point clouds from range-azimuth-intensity measurements and apply point cloud alignment techniques to estimate vehicle motion, scene structure, or to localize within an existing map. In contrast, Dr-BA uses the full radar returns from multiple scans to jointly estimate dense maps and sensor poses. By formulating the problem as a separable optimization, we derive an efficient and general solution that decouples pose estimation from mapping. In addition to solving the BA problem, this formulation naturally extends to direct radar-only localization (DRL) within a previously built map. Dr-BA achieves state-of-the-art radar-based BA and cross-session localization performance, demonstrated on more than 200 km of on-road data across five distinct routes. Our implementation is publicly available at https://github.com/utiasASRL/dr_ba.