Recently, I wrote a piece for Multiple Views: Visualization Research Explained on Learning to Volume Render with GANs. Thanks to Danielle Szafir and Jessica Hullman for inviting me!

This article provides some hindsight on our recent paper A Generative Model for Volume Rendering coauthored by Matt Berger and Jixian Li and published in IEEE TVCG 2019 and presented at IEEE VIS 2018. I attempt to explain the hard bits without going too far into the details.

I also take the opportunity to soapbox a bit on how machine learning is changing the field of visualization.