Deep Learning Portfolio

Art with Deep Learning

Business Case

The client negotiates pictures that she creates from clients’ photos or famous paintings. As both products please customers, she deduced that they would like to have pictures of their family’s in the styles of famous artists and came to us to produce the solution. That insight allowed her to raise the price by 300%.

The Kiss, Gustav Klimt
The Kiss, Gustav Klimt – previous client’s work

Our Solution

We implemented a Deep Learning algorithm called Neural Style Transfer, created by Gatys et al. (2015). We first learned that solution from Andrew Ng deep learning specialization. The algorithm takes three images, a content image, a style reference image (such as an artwork by a famous painter), and the input image you want to style — and blend them together such that the input image is transformed to look like the content image, but like a painting in the style of the style image. At bellow example, it’s a mix of our founder’s dog with a painting from a famous Brazilian artist, Romero Britto.

Content image
+ Style image =
Generated image

At 4tune, we are passionate about solving clients’ problems to deliver profit. If you have a business idea or need help with anything related to data science or artificial intelligence, we are glad to help. Please contact us at

2 replies on “Art with Deep Learning”

Leave a Reply

Your email address will not be published. Required fields are marked *