Research Article
Pillar-Based 3D Object Detection from Point Cloud with Multiattention Mechanism
Table 2
Results on the KITTI test 3D detection benchmark.
| | Model | Car | Pedestrian | Cyclist | | Easy (%) | Mod. (%) | Hard (%) | mAP (%) | Easy (%) | Mod. (%) | Hard (%) | mAP (%) | Easy (%) | Mod. (%) | Hard (%) | mAP (%) |
| | PointPillars | 79.05 | 74.99 | 68.30 | 74.11 | 52.08 | 43.53 | 41.49 | 45.70 | 75.78 | 59.07 | 52.92 | 62.59 | | PointPillars + point-wise attention | 86.82 | 75.80 | 72.94 | 78.52 | 53.01 | 46.74 | 42.96 | 47.57 | 78.71 | 59.84 | 56.56 | 65.03 | | PointPillars + SOPA | 86.13 | 75.27 | 72.86 | 78.08 | 53.44 | 47.61 | 43.45 | 48.10 | 79.34 | 61.63 | 59.18 | 66.72 |
|
|
