Research map / Detection and segmentation / 3D detection and driving
Curricular Object Manipulation in LiDAR-Based Object Detection
Read the paperCheck 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 |
|---|---|---|---|
| Replace GT-Aug with COMLoss + COMAug | SECOND · Waymo validation, Pedestrian | 3D AP_L1 (%): 65.40 -> 66.66 | Helped |
| Replace GT-Aug with COMLoss + COMAug | PointPillars · Waymo validation, Pedestrian | 3D AP_L2 (%): 57.20 -> 58.52 | Helped |
| Replace GT-Aug with COMLoss + COMAug | CenterPoint · Waymo validation, Pedestrian | 3D AP_L1 (%): 73.62 -> 75.43 | Helped |
| Replace GT-Aug with COMLoss + COMAug | SECOND · Waymo validation, Vehicle | 3D AP_L2 (%): 63.95 -> 65.34 | Helped |
| Replace GT-Aug with COMLoss + COMAug | PointPillars · Waymo validation, Vehicle | 3D AP_L1 (%): 70.00 -> 71.38 | Helped |
| Replace GT-Aug with COMLoss + COMAug | CenterPoint · Waymo validation, Vehicle | 3D AP_L1 (%): 71.05 -> 72.15 | Helped |
| Use easy-to-hard COMLoss weighting | CenterPoint + GT-Aug, vehicle · Waymo validation | 3D AP_L1 (%): +0.5% | Helped |
| Use reverse-curricular weighting (hard examples first) (ablation) | CenterPoint + GT-Aug, vehicle · Waymo validation | 3D AP_L1 (%) | Hurt |
| Increase COMLoss curve steepness (ablation) | CenterPoint + GT-Aug, vehicle · Waymo validation | 3D AP_L1 (%) | Hurt |
| Set COMLoss reweighting strength H to 0.6 (ablation) | CenterPoint + GT-Aug, vehicle · Waymo validation | 3D AP_L1 (%): >0.5% | Helped |
| Use distance-only clustering (ablation) | COMAug on CenterPoint, vehicle · Waymo validation | 3D AP_L1 (%): 71.05 -> 71.51 | Helped |
| Add occupancy, angle, and size clustering factors to distance (ablation) | COMAug on CenterPoint, vehicle · Waymo validation | 3D AP_L1 (%): 71.51 -> 72.03 | Helped |
Closest papers by meaning
- A-Teacher: Asymmetric Network for 3D Semi-Supervised Object Detection (CVPR 2024 · similarity 0.87)
- Beyond Virtual Points: Depth-Enhanced LiDAR-only 3D Object Detection with Semi-Supervised Learning (Student Abstract) (AAAI 2025 · similarity 0.86)
- An Empirical Study of the Generalization Ability of Lidar 3D Object Detectors to Unseen Domains (CVPR 2024 · similarity 0.85)
- GBlobs: Explicit Local Structure via Gaussian Blobs for Improved Cross-Domain LiDAR-based 3D Object Detection (CVPR 2025 · similarity 0.85)
- Weak-to-Strong 3D Object Detection with X-Ray Distillation (CVPR 2024 · similarity 0.85)
- Pre-training LiDAR-based 3D Object Detectors through Colorization (ICLR 2024 · similarity 0.85)
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.