AI Paper Atlas

Research map / Detection and segmentation / 3D detection and driving

WeakM3D: Towards Weakly Supervised Monocular 3D Object Detection

Liang Peng, Senbo Yan, Boxi Wu, Zheng Yang, Xiaofei He, Deng Cai

ICLR 2022 · Detection and segmentation › 3D detection and driving · approach: Monocular 3D object detection · 12 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 geometric alignment loss (ablation)WeakM3D with only center loss · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 0.11/0.08 -> 49.09/37.79Improved
Add learning disentanglement (ablation)WeakM3D with only center loss · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 0.11/0.08 -> 16.16/11.55Improved
Add ray tracing to geometric alignment (ablation)WeakM3D with learning disentanglement and geometric alignment · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 51.36/45.04 -> 55.31/48.35Improved
Balance point-wise losses by point density (ablation)WeakM3D with learning disentanglement, geometric alignment, and ray tracing · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 55.31/48.35 -> 58.20/50.16Improved
Learn dimensions rather than freeze them (ablation)WeakM3D · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 58.20/50.16 -> 56.24/48.37Hurt
Use 2D masks rather than 2D boxes to select LiDAR points (ablation)WeakM3D · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 55.94/47.52 -> 58.20/50.16Improved
Use ResNet50 instead of ResNet18 (ablation)WeakM3D · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 50.92/44.78 -> 58.20/50.16Improved
Apply weak supervision rather than full supervision on the basic WeakM3D baselineWeakM3D baseline · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 53.61/44.78 -> 58.20/50.16Improved
Apply weak supervision rather than full supervision with PatchNetPatchNet · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 64.31/57.94 -> 59.41/54.10Hurt
Use WeakM3D instead of M3D-RPNWaymo LEVEL 1 · Waymo, AP_BE V, 30-50mAP_BE V: 3.92 -> 7.73Improved
Use WeakM3D instead of AutolabelsKITTI car weakly supervised detection · KITTI val, car, IoU=0.5, R40, EasyAP_BE V/AP_3D: 50.51/38.31 -> 58.20/50.16Improved

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