Learning task-specific features for 3D pointcloud graph creation
This work addresses a domain-specific bottleneck in 3D pointcloud processing for computer vision applications, offering an incremental improvement over existing graph creation methods.
The paper tackled the problem of creating graphs from 3D pointclouds for use with Graph Neural Networks by proposing a method that learns task-specific features through an MLP transformation and k-NN, outperforming a baseline by 0.3 points in accuracy on ModelNet40.
Processing 3D pointclouds with Deep Learning methods is not an easy task. A common choice is to do so with Graph Neural Networks, but this framework involves the creation of edges between points, which are explicitly not related between them. Historically, naive and handcrafted methods like k Nearest Neighbors (k-NN) or query ball point over xyz features have been proposed, focusing more attention on improving the network than improving the graph. In this work, we propose a more principled way of creating a graph from a 3D pointcloud. Our method is based on performing k-NN over a transformation of the input 3D pointcloud. This transformation is done by an Multi-Later Perceptron (MLP) with learnable parameters that is optimized through backpropagation jointly with the rest of the network. We also introduce a regularization method based on stress minimization, which allows to control how distant is the learnt graph from our baseline: k-NN over xyz space. This framework is tested on ModelNet40, where graphs generated by our network outperformed the baseline by 0.3 points in overall accuracy.