AI Paper Atlas

Research map / Point clouds / Segmentation and detection

Lidar Point Cloud Guided Monocular 3D Object Detection

Liang Peng, Fei Liu, Zhengxu Yu, Senbo Yan, Dan Deng, Zheng Yang, Haifeng Liu, Deng Cai

ECCV 2022 · Point clouds › Segmentation and detection · approach: LiDAR voxel and pillar detectors · 58 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
add LPCG pseudo labelsMonoGRNet · KITTI val Car moderateAP3D R40: 7.57 -> 10.06Improved
add LPCG pseudo labelsM3D-RPN · KITTI val Car moderateAP3D R40: 11.07 -> 19.61Improved
add LPCG pseudo labelsRTM3D · KITTI val Car moderateAP3D R40: 10.06 -> 19.43Improved
add LPCG pseudo labelsGrooMeD-NMS · KITTI val Car moderateAP3D R40: 14.32 -> 20.46Improved
add LPCG pseudo labelsMonoFlex · KITTI test Car moderateAP3D R40: 13.89 -> 17.80Improved
add LPCG pseudo labelsM3D-RPN · KITTI test Car moderateAP3D R40: 9.71 -> 14.82Improved
use only 100 annotations instead of full 3712 for the LiDAR detector (ablation)LPCG M3D-RPN · KITTI val Car moderateAP3D R40: 19.61 -> 15.37Hurt
use 1000 annotations instead of full 3712 (ablation)LPCG M3D-RPN · KITTI val Car moderateAP3D R40: 19.61 -> 19.29No clear effect
train with 10% labeled plus 90% unlabeled LPCG data instead of 100% labeledM3D-RPN · Waymo Level 1 overall3D mAP: 6.42 -> 6.23No clear effect
add LPCG unlabeled data to 10% labeled dataM3D-RPN · Waymo Level 1 overall3D mAP: 4.14 -> 6.23Improved
swap PV-RCNN for PointPillars as pseudo-label LiDAR detector (ablation)LPCG M3D-RPN · KITTI val Car moderateAP3D R40: 19.61 -> 20.12No clear effect
increase pseudo-labeled training samples from 100 to 26057 (ablation)LPCG M3D-RPN · KITTI val Car moderate IoU 0.7AP3D R40: 0.21 -> 19.61Improved

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