AI is enabling content creation like never before. From LLMs dedicated to writing novels to using algorithms to create Drake songs from scratch, artists and creators are entering a new world of opportunities. Using the artistic style of Paul Gauguin, we will use his art style to reimagine images of SpaceX's Dragon initiative via Neural Style Transfer, an algorithm created by Gatys et al. (2015).
- Implement the neural style transfer algorithm
- Generate Gauguin-esque images via the algorithm
- Define a style cost function & a content cost function
- Enable others to experiment with this tool
Neural Style Transfer is a deep learning optimization technique that is gaining popularity amongst content creators. This transfer merges two images: a "content" image (C) and a "style" image (S). Once merged, a "generated" image (G) is created, which combines the "content" of image C with the "style" of image S.
One popular example is a Monet-style depiction of the Louvre. This combines a stock photo of the museum (content image) with the impressionist style of Claude Monet (style image):
The project will use Tensorflow to generate images in the style of Paul Gauguin, a famous French painter known for his impressionist takes on French Polynesia. The final results, using gauguin.jpg and running it on Jupyter, look like this:
Download the iPYNB file here on GitHub. For the pre-trained models and all the items you need to run the tool on your device, visit Google Drive here: https://drive.google.com/drive/folders/1Jp6AuIIqbyRtoTs-MNhaPQO_JQg96JAt?usp=sharing