AI Paper Atlas

Research map / Point clouds / Point cloud classification

Probabilistic Discrepancy Learning for Roadside LiDAR Scene Completion

Xiaogang Wu, Jinchao Hu, Zixian Wang, Dun Liu, BoXiang Cheng, Yiqiang Wu

CVPR 2026 · Point clouds › Point cloud classification · approach: LiDAR point cloud compression and completion

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
add the global sampling discrepancy loss to the diffusion-only baseline (ablation)PDL variant A · V2X-SeqChamfer distance: 0.39 -> 0.37Improved
add the region discrepancy learning loss on top of the global loss (ablation)PDL variant B · V2X-SeqChamfer distance: 0.37 -> 0.34Improved
add PPDM pose correction on top of both losses (ablation)PDL variant C · V2X-SeqChamfer distance: 0.34 -> 0.29Improved
add PPDM pose correction on top of both losses (ablation)PDL variant C · V2X-Seq3D JSD: 0.53 -> 0.48Improved
add PPDM pose correction on top of both losses (ablation)PDL variant C · V2X-Seq (voxel 0.5)Voxel IoU: 54.7 -> 55.9Improved
add PPDM correction to BEVHeight detections (Easy)BEVHeight · V2X-Seq val, Car 3D AP@0.5 Easy3D AP: 77.78 -> 90.32Improved
add PPDM correction to BEVHeight detections (Moderate)BEVHeight · V2X-Seq val, Car 3D AP@0.5 Moderate3D AP: 65.77 -> 86.78Improved
add PPDM correction to BEVHeight detections (Hard)BEVHeight · V2X-Seq val, Car 3D AP@0.5 Hard3D AP: 65.85 -> 84.26Improved
use PDL (pose correction plus diffusion with discrepancy losses) instead of LiDPMdiffusion scene completion · V2X-SeqChamfer distance: 0.35 -> 0.29Improved
use PDL instead of LiDPM on an unseen dataset without fine-tuningdiffusion scene completion · TUMTraf-V2X (cross-dataset)Chamfer distance: 0.34 -> 0.30Improved
use PDL instead of LiDPM on an unseen datasetdiffusion scene completion · TUMTraf-V2X (voxel 0.5)Voxel IoU: 54.9 -> 55.4Improved

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