Dynamic Edge-Conditioned Filters in Convolutional Neural Networks on Graphs

Abstract

A number of problems can be formulated as prediction on graph-structured data. In this work, we generalize the convolution operator from regular grids to arbitrary graphs while avoiding the spectral domain, which allows us to handle graphs of varying size and connectivity. To move beyond a simple diffusion, filter weights are conditioned on the specific edge labels in the neighborhood of a vertex. Together with the proper choice of graph coarsening, we explore constructing deep neural networks for graph classification. In particular, we demonstrate the generality of our formulation in point cloud classification, where we set the new state of the art, and on a graph classification dataset, where we outperform other deep learning approaches. The source code is available at https://github.com/mys007/ecc

Accepted to CVPR 2017; extended version

References in corpus (5)

• Learning Convolutional Neural Networks for Graphs
• Layer Normalization
• Discriminative Embeddings of Latent Variable Models for Structured Data
• Fractional Max-Pooling
• Graph Based Convolutional Neural Network

Cited by in corpus (15)

• Fast Graph Representation Learning with PyTorch Geometric
• Disease Prediction using Graph Convolutional Networks: Application to Autism Spectrum Disorder and Alzheimer's Disease
• Learning Semantic Segmentation of Large-Scale Point Clouds with Random Sampling
• Cross-Scale Internal Graph Neural Network for Image Super-Resolution
• A Fair Comparison of Graph Neural Networks for Graph Classification
• SO-Net: Self-Organizing Network for Point Cloud Analysis
• Adaptive Graph Convolutional Neural Networks
• Graph Convolutional Neural Networks with Node Transition Probability-based Message Passing and DropNode Regularization
• Dynamic Neural Networks: A Survey
• Multi-Objective De Novo Drug Design with Conditional Graph Generative Model
• Implementing graph neural networks with TensorFlow-Keras
• Learning 3D Representations of Molecular Chirality with Invariance to Bond Rotations
• Bayesian Graph Convolutional Neural Networks Using Non-Parametric Graph Learning
• Learning Graph While Training: An Evolving Graph Convolutional Neural Network
• MG-SAGC: A multiscale graph and its self-adaptive graph convolution network for 3D point clouds