Colorization problem is considered.

Colorization example

I have a training set of unordered graphs (images) with varying number of vertices and edges (color regions and adjacency between them, resp.). enter image description here

A fixed number of features is associated with each vertex (color, area, etc.) and with each edge (length of common border, color contrast, etc.).

For input graph (an image to be colored) the trained model is supposed to specify color for each vertex (color region) of the input graph.

Classic regression algorithms (svm, decision trees, etc.) take fixed sized vectors as input and return fixed sized vectors.

Is there a model that naturally handles varying number of vertices / edges?

Is there an approach to adapt fixed structure models (MRF, factor graphs, etc.) to handle this particular case?

  • $\begingroup$ The first image is from «Probabilistic Color-by-Numbers: Suggesting Pattern Colorizations Using Factor Graphs» where authors present a probabilistic factor graph model for automatically coloring 2D patterns. Unfortunately their approach is limited by fixed size of color palette. $\endgroup$ – Ivan Goremykin Sep 22 '15 at 14:32
  • $\begingroup$ Here several papers on graph-based learning are listed. $\endgroup$ – Ivan Goremykin Feb 27 '16 at 21:54
  • $\begingroup$ I moved your post to a comment because it doesn't answer the question. $\endgroup$ – whuber Feb 27 '16 at 23:36
  • $\begingroup$ @whuber No, it does. $\endgroup$ – Ivan Goremykin Feb 28 '16 at 8:42
  • $\begingroup$ According to the standards of this site, a link-only answer is not an answer. Links are fine provided they are accompanied by effective descriptions or summaries of their contents. $\endgroup$ – whuber Feb 28 '16 at 16:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.