Overview / Usage

Neural style transfer has been a fascinating use-case of deep learning. This project showcases the use of automatic differentiation and the Adam optimization technique to update a target image according to content and style images. The aim of this project is to expose the style transfer model via a REST API which can be consumed by the developers that want to use it for their work.

Methodology / Approach

• Custom layer training
• Automatic differentiation
• Custom weights for each of the style layers

Technologies Used

• Python
• TensorFlow 2.0

Repository

https://github.com/sayakpaul/TF-2.0-Hacks/blob/master/Neural_Style_Transfer_(without_TV_loss)_in_TF_2_0.ipynb