HiMo: High-Speed Objects Motion Compensation in Point Clouds
Sina Sharif Mansouri , Olov Andersson , Patric Jensfelt
AI summary
Problem
Mechanical LiDAR sensors suffer from rolling shutter distortions caused by the motion of dynamic objects, which standard ego-motion compensation fails to fix, leading to inaccurate object shapes and positions in high-speed or multi-LiDAR environments.
Approach
HiMo estimates point-wise 3D velocity using self-supervised scene flow models to compute distortion correction vectors, which are then applied to raw point clouds to recover accurate geometry. The pipeline is paired with SeFlow++, a refined real-time scene flow estimator optimized for high-speed scenarios.
Key results
- First pipeline for non-ego motion compensation in raw LiDAR point clouds
- SeFlow++ scene flow estimator achieving state-of-the-art real-time performance
- New Scania dataset capturing high-speed highway driving with multi-LiDAR setups
- Improved geometric consistency and accuracy in semantic segmentation and 3D detection
Why it matters
Essential for autonomous heavy vehicles and high-speed driving systems where precise dynamic object perception directly impacts safety and navigation reliability.
Abstract
LiDAR point cloud is essential for autonomous vehicles, but motion distortions from dynamic objects degrade the data quality. While previous work has considered distortions caused by ego motion, distortions caused by other moving objects remain largely overlooked, leading to errors in object shape and position. This distortion is particularly pronounced in high-speed environments such as highways and in multi-LiDAR configurations, a common setup for heavy vehicles. To address this challenge, we introduce HiMo, a pipeline that repurposes scene flow estimation for non-ego motion compensation, cor- recting the representation of dynamic objects in point clouds. During the development of HiMo, we observed that existing self- supervised scene flow estimators often produce degenerate or inconsistent estimates under high-speed distortion. We further propose SeFlow++, a real-time scene flow estimator that achieves state-of-the-art performance on both scene flow and motion compensation. Since well-established motion distortion metrics are absent in the literature, we introduce two evaluation metrics: compensation accuracy at a point level and shape similarity of objects. We validate HiMo through extensive experiments on Argoverse 2, ZOD and a newly collected real-world dataset featuring highway driving and multi-LiDAR-equipped heavy vehicles. Our findings show that HiMo improves the geometric consistency and visual fidelity of dynamic objects in LiDAR point clouds, benefiting downstream tasks such as semantic segmenta- tion and 3D detection. See https://kin-zhang.github.io/HiMo for more details.