What benefits can we got by applying Graph Convolutional Neural Network instead of ordinary CNN? I mean if we can solve a problem by CNN, what is the reason should we convert to Graph Convolutional Neural Network to solve it? Are there any examples i.e. papers can show by replacing ordinary CNN with Graph Convolutional Neural Network, an accuracy increasement or a quality improvement or a performance gain is achieved? Can anyone introduce some examples as image classification, image recognition especially in medical imaging, bioinfomatics or biomedical areas?
$\begingroup$
$\endgroup$
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
Generally speaking a graph CNN is applied to data represented by graphs, not images.
a graph is a collection of nodes and edges connecting them.
an image is a 2D or 3D matrix, in which each element denotes a pixel in space
If your data are just images, or something similar (e.g. some fMRI data), you usually cannot benefit from graph CNN compared with usual CNN.
Sometimes, the class labels of your images may be organized in a graph-like (or tree-like) structure. In that case, you may have a chance to benefit from graph CNN.
$\begingroup$
$\endgroup$
Bioinformatics is an area that Graph Convolutional Neural Network is useful. Consider protein networks, or gene-gene networks. Surely, the biological networks can be represented as a graph. Now, you should see how GCN is useful for bioinformatics.
