MANet: End-to-End Learning for Point Cloud Based on Robust Pointpillar and Multiattention

<div>Results of 3D detection on the nuScenes test set. The first column is Pointpillar, the second column is our algorithm, and the third is the ground truth.</div>

Wireless Communications and Mobile Computing

fig10

Figure 10

Figure 10: MANet: End-to-End Learning for Point Cloud Based on Robust Pointpillar and Multiattention 