PointNet++ is one of the most influential neural architectures for point cloud understanding.
Although the accuracy of PointNet++ has been largely surpassed by recent networks such …

Transformer, as an alternative to CNN, has been proven effective in many modalities (eg,
texts and images). For 3D point cloud transformers, existing efforts focus primarily on …

Recent advances in 3D point cloud analysis bring a diverse set of network architectures to
the field. However, the lack of a unified framework to interpret those networks makes any …

Point cloud analysis has a wide range of applications in many areas such as computer
vision, robotic manipulation, and autonomous driving. While deep learning has achieved …

MLP-Mixer has newly appeared as a new challenger against the realm of CNNs and
Transformer. Despite its simplicity compared to Transformer, the concept of channel-mixing …

In point cloud analysis, point-based methods have rapidly developed in recent years. These
methods have recently focused on concise MLP structures, such as PointNeXt, which have …

Channel pruning has been widely studied as a prevailing method that effectively reduces
both computational cost and memory footprint of the original network while kee** a …

In this work, we show that it is feasible to perform multiple tasks concurrently on point cloud
with a straightforward yet effective multi-task network. Our framework, Poly-PC, tackles the …

Parsing an unstructured point set into constituent local geometry structures (eg, edges or
surfaces) would be helpful for understanding and representing point clouds. This motivates …

This paper researches the unexplored task—point cloud salient object detection (SOD).
Differing from SOD for images, we find the attention shift of point clouds may provoke …