Dense Soft Weighting for Radar Ego-Velocity Estimation

2026-07-29

AI Summary

This work addresses the limitation of conventional radar systems that rely on constant false alarm rate (CFAR) hard thresholding to convert dense spectrograms into sparse point clouds, which prematurely discards Doppler information from weak echoes and degrades ego-velocity estimation accuracy. To overcome this, the authors propose Dense Soft Weighting—a method that eschews binary detection and instead assigns continuous confidence weights to each range-Doppler cell. Coupled with weighted least squares, this approach enables robust velocity estimation and yields a closed-form expression for velocity covariance, facilitating seamless integration with inertial backends. Notably, the method requires no training data or learned uncertainty models, is transferable across single-chip radar configurations, and operates in real time on embedded platforms. Experiments demonstrate a 31%–45% reduction in average absolute pose error compared to the strongest CFAR baseline across three datasets under identical backend conditions.

Abstract

Sensing ego-velocity estimation is fundamental to state estimation in visually degraded environments, where camera- and LiDAR-based pipelines can become unreliable. Millimetre-wave radar is well suited to these conditions because it provides direct Doppler velocity sensing and remains robust to poor illumination, textureless scenes, and airborne particulates. However, conventional radar ego-velocity pipelines typically apply constant false alarm rate (CFAR) thresholding to convert dense radar spectra into sparse point clouds, prematurely discarding sub-threshold returns that may still retain useful Doppler motion cues. We present Dense Soft Weighting, an analytic radar front-end that maps every range-Doppler cell to a continuous confidence metric rather than enforcing a binary detection threshold. Ego-velocity is then estimated using a deterministic robust weighted least-squares formulation, while the same weighted measurements provide a closed-form, measurement-derived velocity covariance for integration with a shared inertial back-end. The method requires no platform-specific training data or learning-based uncertainty model, supporting transfer across single-chip radar configurations. Across two public datasets and one self-collected dataset, Dense Soft Weighting reduces mean absolute pose error by 31-45% relative to the strongest CFAR point-cloud baseline under an identical inertial back-end, while running in real time on embedded hardware.

Problem

• radar ego-velocity estimation
• CFAR thresholding
• Doppler velocity
• dense radar spectra
• motion cues

Innovation

• Dense Soft Weighting
• radar ego-velocity estimation
• continuous confidence weighting
• robust weighted least-squares
• measurement-derived covariance