# Weakly Supervised Cross-Modal Learning for 4D Radar Scene Flow Estimation

arXiv: https://arxiv.org/abs/2605.18507 · HTML: https://mlsanity.com/paper/2605.18507/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 46 of its 66 references that are in the map.
- First posted on arXiv 18 May 2026.

## Builds on (most cited first, 30 of 46)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation](https://mlsanity.com/paper/1612.00593/) (2016)
- [PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space](https://mlsanity.com/paper/1706.02413/) (2017)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [PointPillars: Fast Encoders for Object Detection from Point Clouds](https://mlsanity.com/paper/1812.05784/) (2018)
- [On the Properties of Neural Machine Translation: Encoder-Decoder Approaches](https://mlsanity.com/paper/1409.1259/) (2014)
- [PointConv: Deep Convolutional Networks on 3D Point Clouds](https://mlsanity.com/paper/1811.07246/) (2018)
- [Segment Everything Everywhere All at Once](https://mlsanity.com/paper/2304.06718/) (2023)
- [YOLOv11: An Overview of the Key Architectural Enhancements](https://mlsanity.com/paper/2410.17725/) (2024)
- [Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution](https://mlsanity.com/paper/2007.16100/) (2020)
- [FlowNet3D: Learning Scene Flow in 3D Point Clouds](https://mlsanity.com/paper/1806.01411/) (2018)
- [FLOT: Scene Flow on Point Clouds Guided by Optimal Transport](https://mlsanity.com/paper/2007.11142/) (2020)
- [Just Go with the Flow: Self-Supervised Scene Flow Estimation](https://mlsanity.com/paper/1912.00497/) (2019)
- [Neural Scene Flow Prior](https://mlsanity.com/paper/2111.01253/) (2021)
- [PV-RAFT: Point-Voxel Correlation Fields for Scene Flow Estimation of Point Clouds](https://mlsanity.com/paper/2012.00987/) (2020)
- [FlowStep3D: Model Unrolling for Self-Supervised Scene Flow Estimation](https://mlsanity.com/paper/2011.10147/) (2020)
- [CamLiFlow: Bidirectional Camera-LiDAR Fusion for Joint Optical Flow and Scene Flow Estimation](https://mlsanity.com/paper/2111.10502/) (2021)
- [Scalable Scene Flow from Point Clouds in the Real World](https://mlsanity.com/paper/2103.01306/) (2021)
- [Scene Flow from Point Clouds with or without Learning](https://mlsanity.com/paper/2011.00320/) (2020)
- [Fast Neural Scene Flow](https://mlsanity.com/paper/2304.09121/) (2023)
- [SeFlow: A Self-Supervised Scene Flow Method in Autonomous Driving](https://mlsanity.com/paper/2407.01702/) (2024)
- [Bi-PointFlowNet: Bidirectional Learning for Point Cloud Based Scene Flow Estimation](https://mlsanity.com/paper/2207.07522/) (2022)
- [Hidden Gems: 4D Radar Scene Flow Learning Using Cross-Modal Supervision](https://mlsanity.com/paper/2303.00462/) (2023)
- [ICP-Flow: LiDAR Scene Flow Estimation with ICP](https://mlsanity.com/paper/2402.17351/) (2024)
- [DeFlow: Decoder of Scene Flow Network in Autonomous Driving](https://mlsanity.com/paper/2401.16122/) (2024)
- [Self-Supervised Scene Flow Estimation with 4-D Automotive Radar](https://mlsanity.com/paper/2203.01137/) (2022)
- [RPEFlow: Multimodal Fusion of RGB-PointCloud-Event for Joint Optical Flow and Scene Flow Estimation](https://mlsanity.com/paper/2309.15082/) (2023)
- [Flow4D: Leveraging 4D Voxel Network for LiDAR Scene Flow Estimation](https://mlsanity.com/paper/2407.07995/) (2024)

## Abstract (arXiv)

Due to the difficulty of obtaining ground-truth data for 4D radar scene flow estimation, previous methods typically rely on either self-supervised losses or cross-modal supervision using 3D LiDAR data, 2D images, and odometry. However, self-supervised approaches often yield suboptimal results due to radar's inherently low-fidelity measurements, while existing cross-modal supervised methods introduce complex multi-task architecture and require costly LiDAR sensors to generate pseudo radar scene flow labels from pretrained 3D tracking models. To overcome these limitations, we propose a task-specific iterative framework for weakly supervised radar scene flow learning, using only images and odometry for auxiliary supervision during training. Specially, we establish two novel instance-aware self-supervised losses by exploiting off-the-shelf 2D tracking and segmentation algorithms to obtain tracked instance masks, which are back-projected into 3D space to provide instance-level semantic guidance; for static regions, we integrate vehicle odometry with radar's intrinsic motion cues to construct a rigid static loss. Extensive experiments on the real-world View-of-Delft (VoD) dataset demonstrate that our method not only surpasses state-of-the-art cross-modal supervised approaches that rely on 3D multi-object tracking on dense LiDAR point clouds but also outperforms existing fully supervised scene flow estimation methods. The code is open-sourced at hrefhttps://github.com/FuJingyun/IterFlowhttps://github.com/FuJingyun/IterFlow.