Generate Art NFTs from Neural Style Transfer method by help of Eager Execution
This project helps designers and painters put their arts on the showcase and other users can mint NFT of their choice from these art styles by paying artists in return. This way a user can have an NFT of his choice by merging the style from the painter's art and the image of his choice.
How it's made
On the backend, it utilizes TensorFlow and Matplotlib for the generation of Images and Flask for HTTP requests. On the Frontend side, the Next Js application is there with web3 js for user state management and NFT Port API for minting NFTs. The matplotlib breaks down the images in gram matrix and calculates loss then net loss of style and the content image is calculated into another gram matrix to obtain the final image