360 Style Transfers

For my Thesis project at ITP, I figured out a way to correctly apply style transfers to 360 imagery. This was a continuation of an earlier ITP project, Traveling in Style with Google Street View.

Here is my Thesis Statement:

I designed and implemented a new approach for applying style transfers to 360 pictures and video. Using contemporary and historical paintings and images from Google Street View, I created art that encourages viewers to appreciate what is beautiful or poignant about the modern world, and to see it with new eyes.

Below is an example of my work. The content image is from Bessy, France, and was obtained using the Google Street View API. The style image is a water color painting I made in a class as part of my thesis research. The main idea here is to show people the world through the eyes of an artist instead of the mechanical eye of Google's camera.

You can see more examples of my work at the website I created for this project, 360.ixora.io. If you are interested, my Thesis presentation video is also available:

Eventually I will document my style transfer algorithm in a paper I will submit for publication. I will also release an open source library to share with the community. I hope to have this done by the end of 2019.