AI Paper Atlas

Research map / Point clouds / Point cloud classification

UniFlow: Zero-Shot LiDAR Scene Flow for Autonomous Driving

Siyi Li, Qingwen Zhang, Ishan Khatri, Kyle Vedder, Eric Eaton, Deva Ramanan, Neehar Peri

ECCV 2026 · Point clouds › Point cloud classification · approach: LiDAR point cloud compression and completion · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train on unified multi-dataset mixture (AV2, Waymo, nuScenes) instead of Waymo only (ablation)Flow4D · TruckScenes zero-shotDynamic Mean EPE: 0.336 -> 0.310Improved
train on unified multi-dataset mixture instead of Waymo only (ablation)SSF · TruckScenes zero-shotDynamic Mean EPE: 0.737 -> 0.577Improved
train on unified multi-dataset mixture instead of Waymo only (ablation)DeltaFlow · TruckScenes zero-shotDynamic Mean EPE: 0.357 -> 0.331Improved
add height randomization and LiDAR ray dropping augmentation (ablation)DeltaFlow (UniFlow) · TruckScenes zero-shotDynamic Mean EPE: 0.331 -> 0.283Improved
add augmentation to unified training (ablation)SSF (UniFlow) · TruckScenes zero-shotDynamic Mean EPE: 0.577 -> 0.435Improved
add augmentation to unified training (ablation)Flow4D (UniFlow) · TruckScenes zero-shotDynamic Mean EPE: 0.310 -> 0.301Improved
scale up Flow4D backbone to XL (ablation)Flow4D (UniFlow) · TruckScenes zero-shotDynamic Mean EPE: 0.301 -> 0.281Improved
multi-dataset UniFlow training instead of dataset-specific trainingFlow4D-XL · Argoverse 2 testDynamic Mean EPE: 0.145 -> 0.132Improved
multi-dataset UniFlow training instead of dataset-specific trainingSSF · nuScenes valDynamic Mean EPE: 0.220 -> 0.144Improved
multi-dataset UniFlow training instead of dataset-specific trainingDeltaFlow · Argoverse 2 testDynamic Mean EPE: 0.113 -> 0.118No clear effect
multi-dataset UniFlow training instead of best single-dataset modelDeltaFlow · AEVAScenes zero-shotDynamic Mean EPE: 0.391 -> 0.303Improved
multi-dataset UniFlow training instead of AV2-only for long rangeDeltaFlow-LR · AV2 long-rangeDynamic Mean EPE: 0.259 -> 0.226Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy