diffConv: Analyzing Irregular Point Clouds with an Irregular View
Standard spatial convolutions assume input data with a regular neighborhood structure. Existing methods typically generalize convolution to the irregular point cloud domain by fixing a regular "view" through e.g. a fixed neighborhood size, where the convolution kernel size remains the same for each point. However, since point clouds are not as structured as images, the fixed neighbor number gives an unfortunate inductive bias. We present a novel graph convolution named Difference Graph Convolution (diffConv), which does not rely on a regular view. diffConv operates on spatially-varying and density-dilated neighborhoods, which are further adapted by a learned masked attention mechanism. We validate our model on the ModelNet40 point cloud classification benchmark, obtaining state-of-the-art performance and far more robustness to noise, along with a much faster inference speed.
READ FULL TEXT