Research map / Point clouds / Segmentation and detection
OccAM's Laser: Occlusion-Based Attribution Maps for 3D Object Detectors on LiDAR Data
David Schinagl, Georg Krispel, Horst Possegger, Peter M. Roth, Horst Bischof
Check 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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use density-aware, distance-dependent voxel keep probabilities instead of a fixed keep probability (ablation) | OccAM · KITTI, objects at different distances from the LiDAR sensor | mean similarity score across object distances | Improved |
| Use a fixed voxel keep probability instead of density-aware sampling (ablation) | OccAM · KITTI, objects at different distances from the LiDAR sensor | mean similarity score across object distances | Hurt |
| Remove reflectivity values from PointPillars training (ablation) | PointPillars · KITTI car detections | importance of license plates in attribution maps | Hurt |
| Increase the attribution-map sampling iterations (ablation) | OccAM with PointPillars · KITTI individual attribution maps | attribution-map quality: N=100; N=300; N=3000 | Improved |
| Enlarge evaluation bounding boxes by 10% (ablation) | OccAM with PointPillars · KITTI validation | pointing game score: 0.9004 -> 0.9724 | Improved |
| Remove the most important points before other points (ablation) | OccAM attribution maps · KITTI detections | model degradation under point dropping | Improved |
| Remove the least important points before other points (ablation) | OccAM attribution maps · KITTI detections | model degradation under point dropping | Hurt |
Closest papers by meaning
- Not All Points Are Equal: Learning Highly Efficient Point-Based Detectors for 3D LiDAR Point Clouds (CVPR 2022 · similarity 0.90)
- Open-CRB: Toward Open World Active Learning for 3D Object Detection (TPAMI 2025 · similarity 0.89)
- Robust 3D Object Detection using Probabilistic Point Clouds from Single-Photon LiDARs (ICCV 2025 · similarity 0.89)
- OpenSight: A Simple Open-Vocabulary Framework for LiDAR-Based Object Detection (ECCV 2024 · similarity 0.89)
- GBlobs: Explicit Local Structure via Gaussian Blobs for Improved Cross-Domain LiDAR-based 3D Object Detection (CVPR 2025 · similarity 0.88)
- An Empirical Study of the Generalization Ability of Lidar 3D Object Detectors to Unseen Domains (CVPR 2024 · similarity 0.88)
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.
