AI Paper Atlas

Research map / Point clouds / Segmentation and detection

LMD: Light-Weight Prediction Quality Estimation for Object Detection in Lidar Point Clouds

Tobias Riedlinger, Marius Schubert, Sarina Penquitt, Jan-Marcel Kezmann, Pascal Colling, Karsten Kahl, Lutz Roese-Koerner, Michael Arnold, Urs Zimmermann, Matthias Rottmann

IJCV 2025 · Point clouds › Segmentation and detection · approach: LiDAR voxel and pillar detectors · 2 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
use all 90 LMD features instead of score onlygradient boosting meta classifier · nuScenes CenterPointAUROC: 0.8623 -> 0.9628Improved
add Lidar and proposal statistics beyond box featuresgradient boosting meta classifier · nuScenes CenterPointAUROC: 0.9529 -> 0.9628Improved
add box features to scoregradient boosting meta regressor · nuScenes CenterPointR^2: 0.4732 -> 0.6792Improved
add all LMD features to box featuresgradient boosting meta regressor · nuScenes PointPillarsR^2: 0.5593 -> 0.6413Improved
add all LMD features to box featuresgradient boosting meta regressor · KITTI PointPillarsR^2: 0.7131 -> 0.7287Improved
use non-linear meta classifier (random forest) instead of logistic regression (ablation)LMD features · nuScenes CenterPointAUROC: 0.9450 -> 0.9581Improved
use gradient boosting instead of logistic regression (ablation)LMD features · nuScenes CenterPointAUROC: 0.9450 -> 0.9628Improved
use meta model choice on score only (ablation)score baseline · nuScenes CenterPointAUROC: 0.8644 -> 0.8617No clear effect
use gradient boosting instead of ridge regression (ablation)LMD features · nuScenes CenterPointR^2: 0.6451 -> 0.7296Improved
select only five greedy features instead of all 90 (ablation)LMD · nuScenes CenterPointAUROC: 0.19 pp worseNo clear effect
meta classifier confidence instead of objectness scoredetector confidence · nuScenes CenterPointECE: 8.07 ppImproved
rank false positives by LMD predicted IoU instead of scoreannotation error proposals · nuScenesannotation errors in top 100: 6 -> 43Improved

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