Research map / Detection and segmentation / 3D detection and driving
What You See is What You Get: Exploiting Visibility for 3D Object Detection
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Add a visibility input stream to PointPillars | PointPillars · NuScenes test | mAP: 30.5 -> 35.0 | Helped |
| Add a visibility input stream to PointPillars | PointPillars† · NuScenes validation | mAP: 31.5 -> 35.4 | Helped |
| Use early rather than late fusion for visibility (ablation) | Two-stream PointPillars with drilling and multi-frame aggregation · NuScenes validation | mAP: 34.0 -> 35.4 | Helped |
| Use drilling rather than culling for object augmentation (ablation) | Two-stream PointPillars with late fusion and multi-frame aggregation · NuScenes validation | bus mAP: 11.4% | Helped |
| Use drilling rather than naive object augmentation (ablation) | Two-stream PointPillars with late fusion and multi-frame aggregation · NuScenes validation | bus mAP: 1.9% | Helped |
| Add object augmentation (ablation) | Two-stream PointPillars with late fusion and multi-frame aggregation · NuScenes validation | bus mAP: 13.7% | Helped |
| Add temporal aggregation (ablation) | Two-stream PointPillars with late fusion and no object augmentation · NuScenes validation | mAP: 9.4% | Helped |
| Add the visibility stream (ablation) | Vanilla two-stream PointPillars, single-frame · NuScenes validation | mAP: 1.4% | Helped |
| Use both point-cloud and visibility streams rather than visibility alone (ablation) | PointPillars, single-frame · NuScenes validation | mAP: 4% | Helped |
| Add object augmentation (ablation) | Vanilla PointPillars, single-frame · NuScenes validation | mAP on augmented classes: +9.1% | Helped |
| Add temporal aggregation (ablation) | Vanilla PointPillars · NuScenes validation | mAP: +8.6% | Helped |
| Add object augmentation and temporal aggregation (ablation) | Vanilla PointPillars · NuScenes validation | mAP: +11.0% | Helped |
Closest papers by meaning
- VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection (CVPR 2018 · similarity 0.90)
- VoxelNeXt: Fully Sparse VoxelNet for 3D Object Detection and Tracking (CVPR 2023 · similarity 0.89)
- Point Density-Aware Voxels for LiDAR 3D Object Detection (CVPR 2022 · similarity 0.89)
- JPV-Net: Joint Point-Voxel Representations for Accurate 3D Object Detection (AAAI 2022 · similarity 0.88)
- Voxel or Pillar: Exploring Efficient Point Cloud Representation for 3D Object Detection (AAAI 2024 · similarity 0.88)
- An Empirical Study of the Generalization Ability of Lidar 3D Object Detectors to Unseen Domains (CVPR 2024 · similarity 0.87)
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.