Survey of Image Based Graph Neural Networks

TL;DR

This survey reviews image-based graph neural networks, highlighting a three-step classification approach involving superpixel conversion, graph generation, and GCN processing, with spectral models showing superior performance.

Contribution

The paper provides a comprehensive analysis of image-based GNNs and introduces a novel classification pipeline utilizing superpixels and graph convolutional networks.

Findings

Spectral-based GNN models outperform spatial-based models.

Superpixel conversion reduces input data by 30%.

Spectral models require less computational cost.

Abstract

In this survey paper, we analyze image based graph neural networks and propose a three-step classification approach. We first convert the image into superpixels using the Quickshift algorithm so as to reduce 30% of the input data. The superpixels are subsequently used to generate a region adjacency graph. Finally, the graph is passed through a state-of-art graph convolutional neural network to get classification scores. We also analyze the spatial and spectral convolution filtering techniques in graph neural networks. Spectral-based models perform better than spatial-based models and classical CNN with lesser compute cost.