Today I want to talk about Neural Style Transfer and Convolutional Neural Networks (CNNs). There are already quite a few articles and tutorials available. Sometimes content is just copied, some provide a novel implementation. What all have in common is a very fast dive into specifics. Too specific in my opinion. Not only that, but there are often implementation details that make it harder to focus on the main concept as a whole.
This article can be considered as an overview and comprehension of other articles (listed in my “Inspiration” section), to understand the concept on a higher level. My intention is to strip away some implementation details, being high level enough for beginners and sparking curiosity for reading the original research paper and subsequent implementations.
Table of Contents
- Disclaimer
- Requirements
- Neural Style Transfer
- Architectural and process overview
- Transfer learning and style transfer
- 1. Transfer learning
- 2. Style Transfer
- Style and content
- Cost calculation
- Content cost
- Style cost
- Total variation cost
- Inspiration
- About
Disclaimer
I am not associated with any of the services I use in this article.
I do not consider myself an expert. If you have the feeling that I am missing important steps or neglected something, consider pointing it out in the comment section or get in touch with me.
I am always happy for constructive input and how to improve.
This was written on 2020–10–17. I cannot monitor all of my articles. There is a high probability that when you read this article the tips are outdated and the processes have changed.
If you need more information on certain parts, feel free to point it out in the comments.
#deep-learning #neural-style-transfer #machine-learning
