Pointwise Convolutional Neural Networks
This addresses the problem of 3D data processing for computer vision researchers, presenting an incremental improvement over existing methods.
The paper tackles the challenge of applying convolutional neural networks to 3D point clouds by introducing pointwise convolution, a new operator that processes each point individually. This fully convolutional network achieves competitive accuracy in semantic segmentation and object recognition tasks.
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.