DensePoint: Learning Densely Contextual Representation for Efficient Point Cloud Processing

TL;DR

DensePoint introduces a novel architecture that learns densely contextual representations for point cloud processing, extending grid CNNs to irregular points and aggregating multi-scale semantics for improved performance.

Contribution

It generalizes convolution to irregular point configurations and employs dense connections to enhance contextual semantic learning in point clouds.

Findings

Achieves state-of-the-art results on multiple benchmarks

Effectively captures multi-scale semantic information

Demonstrates high efficiency in point cloud tasks

Abstract

Point cloud processing is very challenging, as the diverse shapes formed by irregular points are often indistinguishable. A thorough grasp of the elusive shape requires sufficiently contextual semantic information, yet few works devote to this. Here we propose DensePoint, a general architecture to learn densely contextual representation for point cloud processing. Technically, it extends regular grid CNN to irregular point configuration by generalizing a convolution operator, which holds the permutation invariance of points, and achieves efficient inductive learning of local patterns. Architecturally, it finds inspiration from dense connection mode, to repeatedly aggregate multi-level and multi-scale semantics in a deep hierarchy. As a result, densely contextual information along with rich semantics, can be acquired by DensePoint in an organic manner, making it highly effective. Extensive experiments on challenging benchmarks across four tasks, as well as thorough model analysis, verify DensePoint achieves the state of the arts.