Table 4 Aggregated AP and mAP across 3 folds × 3 seeds (mean ± std).
From: Cross-dataset late fusion of Camera–LiDAR and radar models for object detection
Class | KITTI AP (Camera + LiDAR) | Radar AP | Fusion AP (vs. KITTI GT) | Fusion AP (vs. Radar GT) |
|---|---|---|---|---|
0 – Car | 0.9983 ± 0.0017 | 0.6860 ± 0.0149 | 0.9974 ± 0.0018 | 0.6829 ± 0.0148 |
1 – Truck/Large Vehicle | 0.9351 ± 0.0100 | 0.2811 ± 0.0196 | 0.9337 ± 0.0115 | 0.2842 ± 0.0110 |
2 – Train/Tram | — | 0.0495 ± 0.0102 | — | 0.0450 ± 0.0045 |
3 – Bicycle | — | — | — | — |
4 – Pedestrian | 0.9268 ± 0.0173 | — | 0.9180 ± 0.0161 | — |
