Published on Thu Feb 22 2018

SPLATNet: Sparse Lattice Networks for Point Cloud Processing

Hang Su, Varun Jampani, Deqing Sun, Subhransu Maji, Evangelos Kalogerakis, Ming-Hsuan Yang, Jan Kautz

We present a network architecture for processing point clouds that directly operates on a collection of points. Naively applying convolutions on this lattice scales poorly, both in terms of memory and computational cost. Instead, our network uses sparse bilateral convolutional layers as building blocks.

0
0
0
Abstract

We present a network architecture for processing point clouds that directly operates on a collection of points represented as a sparse set of samples in a high-dimensional lattice. Naively applying convolutions on this lattice scales poorly, both in terms of memory and computational cost, as the size of the lattice increases. Instead, our network uses sparse bilateral convolutional layers as building blocks. These layers maintain efficiency by using indexing structures to apply convolutions only on occupied parts of the lattice, and allow flexible specifications of the lattice structure enabling hierarchical and spatially-aware feature learning, as well as joint 2D-3D reasoning. Both point-based and image-based representations can be easily incorporated in a network with such layers and the resulting model can be trained in an end-to-end manner. We present results on 3D segmentation tasks where our approach outperforms existing state-of-the-art techniques.

Sat Nov 17 2018
Computer Vision
PointConv: Deep Convolutional Networks on 3D Point Clouds
3D point clouds are irregular and unordered, hence applying convolution on them can be difficult. PointConv can be applied on point clouds to build deep convolutional networks.
0
0
0
Tue Aug 21 2018
Computer Vision
Fully-Convolutional Point Networks for Large-Scale Point Clouds
This work proposes a general-purpose, fully-convolutional network for efficiently processing large-scale 3D data. The network can process unorganized 3Drepresentations such as point clouds as input, then transforming them internally to ordered structures to be processed via 3D convolutions.
0
0
0
Fri Mar 30 2018
Computer Vision
SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filters
Deep neural networks have enjoyed remarkable success for various vision tasks. However it remains challenging to apply CNNs to domains lacking aregular underlying structures. We propose a novel convolutional architecture, termed SpiderCNN, to efficiently extract features from point clouds.
0
0
0
Tue Apr 16 2019
Computer Vision
A-CNN: Annularly Convolutional Neural Networks on Point Clouds
Analyzing the geometric and semantic properties of 3D point clouds through. the deep networks is still challenging due to the irregularity and sparsity of their geometric structures. This paper presents a new method to define and compute convolution directly on 3D. point clouds.
0
0
0
Fri Dec 02 2016
Computer Vision
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
Point cloud is an important type of geometric data structure. Most researchers transform such data to regular 3D voxel grids or collections of images. This, however, renders data unnecessarily voluminous. In this paper, we design a novel type of neural network that directly consumes point clouds.
1
0
1
Mon Nov 25 2019
Machine Learning
RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds
We study the problem of efficient semantic segmentation for large-scale 3D point clouds. The key to our approach is to use random point sampling instead of more complex point selection approaches. Extensive experiments show that our RandLA-Net can process 1 million points in a single pass.
1
3
11