UniFlow: Towards Zero-Shot LiDAR Scene Flow for Autonomous Vehicles via Cross-Domain Generalization

• URL: http://arxiv.org/abs/2511.18254v1
• Date: Sun, 23 Nov 2025 02:51:42 GMT
• Title: UniFlow: Towards Zero-Shot LiDAR Scene Flow for Autonomous Vehicles via Cross-Domain Generalization
• Authors: Siyi Li, Qingwen Zhang, Ishan Khatri, Kyle Vedder, Deva Ramanan, Neehar Peri,
• Abstract summary: We learn general motion priors that transfer to diverse and unseen LiDAR sensors.<n>We propose UniFlow, a family of feedforward models that unifies and trains on multiple large-scale LiDAR scene flow datasets.<n>UniFlow achieves state-of-the-art accuracy on unseen datasets like TruckScenes, outperforming prior TruckScenes-specific models by 30.1%.
• Score: 46.6187296172395
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: LiDAR scene flow is the task of estimating per-point 3D motion between consecutive point clouds. Recent methods achieve centimeter-level accuracy on popular autonomous vehicle (AV) datasets, but are typically only trained and evaluated on a single sensor. In this paper, we aim to learn general motion priors that transfer to diverse and unseen LiDAR sensors. However, prior work in LiDAR semantic segmentation and 3D object detection demonstrate that naively training on multiple datasets yields worse performance than single dataset models. Interestingly, we find that this conventional wisdom does not hold for motion estimation, and that state-of-the-art scene flow methods greatly benefit from cross-dataset training. We posit that low-level tasks such as motion estimation may be less sensitive to sensor configuration; indeed, our analysis shows that models trained on fast-moving objects (e.g., from highway datasets) perform well on fast-moving objects, even across different datasets. Informed by our analysis, we propose UniFlow, a family of feedforward models that unifies and trains on multiple large-scale LiDAR scene flow datasets with diverse sensor placements and point cloud densities. Our frustratingly simple solution establishes a new state-of-the-art on Waymo and nuScenes, improving over prior work by 5.1% and 35.2% respectively. Moreover, UniFlow achieves state-of-the-art accuracy on unseen datasets like TruckScenes, outperforming prior TruckScenes-specific models by 30.1%.

Related papers

• Learning to Track Any Points from Human Motion [55.831218129679144]
  We propose an automated pipeline to generate pseudo-labeled training data for point tracking.<n>A point tracking model trained on AnthroTAP achieves annotated state-of-the-art performance on the TAP-Vid benchmark.
  arXiv  Detail & Related papers  (2025-07-08T17:59:58Z)
• Unsupervised Domain Adaptation for Self-Driving from Past Traversal Features [69.47588461101925]
  We propose a method to adapt 3D object detectors to new driving environments. Our approach enhances LiDAR-based detection models using spatial quantized historical features. Experiments on real-world datasets demonstrate significant improvements.
  arXiv  Detail & Related papers  (2023-09-21T15:00:31Z)
• SPOT: Scalable 3D Pre-training via Occupancy Prediction for Learning Transferable 3D Representations [76.45009891152178]
  Pretraining-finetuning approach can alleviate the labeling burden by fine-tuning a pre-trained backbone across various downstream datasets as well as tasks.<n>We show, for the first time, that general representations learning can be achieved through the task of occupancy prediction.<n>Our findings will facilitate the understanding of LiDAR points and pave the way for future advancements in LiDAR pre-training.
  arXiv  Detail & Related papers  (2023-09-19T11:13:01Z)
• InsMOS: Instance-Aware Moving Object Segmentation in LiDAR Data [13.196031553445117]
  We propose a novel network that addresses the challenge of segmenting moving objects in 3D LiDAR scans. Our method exploits a sequence of point clouds as input and quantifies them into 4D voxels. We use 4D sparse convolutions to extract motion features from the 4D voxels and inject them into the current scan.
  arXiv  Detail & Related papers  (2023-03-07T14:12:52Z)
• A Lightweight and Detector-free 3D Single Object Tracker on Point Clouds [50.54083964183614]
  It is non-trivial to perform accurate target-specific detection since the point cloud of objects in raw LiDAR scans is usually sparse and incomplete. We propose DMT, a Detector-free Motion prediction based 3D Tracking network that totally removes the usage of complicated 3D detectors.
  arXiv  Detail & Related papers  (2022-03-08T17:49:07Z)
• FAST3D: Flow-Aware Self-Training for 3D Object Detectors [12.511087244102036]
  State-of-the-art self-training approaches mostly ignore the temporal nature of autonomous driving data. We propose a flow-aware self-training method that enables unsupervised domain adaptation for 3D object detectors on continuous LiDAR point clouds. Our results show a significant improvement over the state-of-the-art, without any prior target domain knowledge.
  arXiv  Detail & Related papers  (2021-10-18T14:32:05Z)
• Scalable Scene Flow from Point Clouds in the Real World [30.437100097997245]
  We introduce a new large scale benchmark for scene flow based on the Open dataset. We show how previous works were bounded based on the amount of real LiDAR data available. We introduce the model architecture FastFlow3D that provides real time inference on the full point cloud.
  arXiv  Detail & Related papers  (2021-03-01T20:56:05Z)
• 2nd Place Scheme on Action Recognition Track of ECCV 2020 VIPriors Challenges: An Efficient Optical Flow Stream Guided Framework [57.847010327319964]
  We propose a data-efficient framework that can train the model from scratch on small datasets. Specifically, by introducing a 3D central difference convolution operation, we proposed a novel C3D neural network-based two-stream framework. It is proved that our method can achieve a promising result even without a pre-trained model on large scale datasets.
  arXiv  Detail & Related papers  (2020-08-10T09:50:28Z)
• InfoFocus: 3D Object Detection for Autonomous Driving with Dynamic Information Modeling [65.47126868838836]
  We propose a novel 3D object detection framework with dynamic information modeling. Coarse predictions are generated in the first stage via a voxel-based region proposal network. Experiments are conducted on the large-scale nuScenes 3D detection benchmark.
  arXiv  Detail & Related papers  (2020-07-16T18:27:08Z)
• High-Precision Digital Traffic Recording with Multi-LiDAR Infrastructure Sensor Setups [0.0]
  We investigate the impact of fused LiDAR point clouds compared to single LiDAR point clouds. The evaluation of the extracted trajectories shows that a fused infrastructure approach significantly increases the tracking results and reaches accuracies within a few centimeters.
  arXiv  Detail & Related papers  (2020-06-22T10:57:52Z)
• Any Motion Detector: Learning Class-agnostic Scene Dynamics from a Sequence of LiDAR Point Clouds [4.640835690336654]
  We propose a novel real-time approach of temporal context aggregation for motion detection and motion parameters estimation. We introduce an ego-motion compensation layer to achieve real-time inference with performance comparable to a naive odometric transform of the original point cloud sequence.
  arXiv  Detail & Related papers  (2020-04-24T10:40:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.