Scale-Aware Attention-Based PillarsNet (SAPN) Based 3D Object Detection for Point Cloud

Table 5

Average precisions of 3D detection on the KITTI validation split. , , and represent average precisions of easy, moderate, and hard splits, respectively.

Multi-resolution	Spatial-attention	SE-attention
✓	✕	✕	86.12	76.12	68.56
✕	✓	✕	86.47	77.22	68.86
✕	✕	✓	86.01	76.03	68.13
✓	✓	✕	87.86	77.86	69.86
✓	✕	✓	86.34	76.88	68.78
✕	✓	✓	86.65	77.98	69.12
✓	✓	✓	88.38	78.28	69.52