… small-scale point clouds. In this paper, we introduce RandLA-Net, an efficient and lightweight
neural architecture to directly infer per-point semantics for large-scale point clouds. The …

… small-scale point clouds. In this paper, we introduce RandLA-Net, an efficient and lightweight
neural architecture to directly infer per- point semantics for large-scale point clouds. The …

… problem of 3D point cloud semantic segmentation for large-scale … structure of RandLA-Net,
a deeper 3D point cloud semantic … Based on the RandLA-Net algorithm, we replaced random …

… To further demonstrate the advantages of our method compared with RandLA-Net, more
results have been given in Table 2. In this table, we have listed the mIOU of segmentation …

… largescale urban point clouds and an outlook on future directions for 3D point cloud segmentation
at massive scale … baselines PointNet [37] and RandLA-Net [23], we evaluate how the …

… for point cloud segmentation, namely, RandLA-Net, an … and decoder features of the same
scales. Second, a channel attentive … of RandLA-Net [23], which sampled the whole point cloud …

… In Table 4, we use a similar metric to the RandLA-Net to evaluate the performance of our
method and the mainstream methods. Compared to the RandLA-Net, our FARnet requires …

… 2, we propose a novel method named LG-Net for large-scale point cloud semantic segmentation.
… Compared with RandLA-Net and GA-Net, our method improves OA by 1.0% and 0.5%, …

… Moreover, we implement RandLA-Net on the selfsupervised task by modifying the final
output layer. That is, the output of the network is a 6-dimension vector which contains the …

… Net thanks to these two modules and evaluate LGS-Net on three public large-scale point
clouds … Compared with RandLA-Net, we design the LGSR module to represent z-axis rotation-…