Skip to main content
. 2022 Mar 14;17(3):e0263916. doi: 10.1371/journal.pone.0263916

Table 2. Performance results of PointNet++.

Evaluation Mean Loss 0.0613 ± 0.0477
Evaluation Accuracy 0.9764 ± 0.0210
Evaluation Average Class Accuracy 0.9517 ± 0.04751
Mean Intersection over Union (MIoU) 0.9220 ± 0.05630