Artistic Style Transfer (using Deep Learning)
- Vincent Toscano

- Aug 9, 2016
- 1 min read
Updated: Apr 9, 2020

I was curious about how researchers were using Deep Neural Networks in the context of creative work. When I discovered a paper on Artistic Style Transfers, I knew I had to give it a try. The above image is my attempt at applying the artistic style of Claude Monet to a wonderfully composited image from the the magicians at ILM for the movie "The Phantom Menace" (a film I spent a number of years working on in R&D.)
To achieve the above image the software implementation uses neural representations to separate and recombine content and style of arbitrary images. The implementation is from this gentleman's work: artistic-style-transfer
He provides an implementation based on a paper from Cornell and using the libraries from Berkeley's Vision and Learning Center. The inputs I provide (from right to left) would be the style and content source images, with the last image being the output from the technique.







Comments