Pointwise Convolutional Neural Networks
arXiv:1712.05245
Abstract
Deep learning with 3D data such as reconstructed point clouds and CAD models has received great research interests recently. However, the capability of using point clouds with convolutional neural network has been so far not fully explored. In this paper, we present a convolutional neural network for semantic segmentation and object recognition with 3D point clouds. At the core of our network is pointwise convolution, a new convolution operator that can be applied at each point of a point cloud. Our fully convolutional network design, while being surprisingly simple to implement, can yield competitive accuracy in both semantic segmentation and object recognition task.
arXiv comment10 pages, 6 figures, 10 tables. Paper accepted to CVPR 2018
References in corpus (13)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Deep Residual Learning for Image Recognition
- Geometric deep learning: going beyond Euclidean data
- Multi-Scale Context Aggregation by Dilated Convolutions
- Fully Convolutional Networks for Semantic Segmentation
- DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
- CNN Features off-the-shelf: an Astounding Baseline for Recognition
- Rich feature hierarchies for accurate object detection and semantic segmentation
- Multi-view Convolutional Neural Networks for 3D Shape Recognition
- PointCNN: Convolution On -Transformed Points
- PU-Net: Point Cloud Upsampling Network
- SemanticFusion: Dense 3D Semantic Mapping with Convolutional Neural Networks
- Semantic Scene Completion from a Single Depth Image