Neural Styled NFT Minter can generate and deploy NFTs on blockchain using Deep Learning based Neural Style Transfer to enable computer generated creative art pieces
- Frontend: node.js, yarn
- Backend: Python libraries:
tensorflow tensorflow_hub flask flask_cors Pillow
- Other services used: NFT.STORAGE, InterPlanetary File System, hardhat,
-
Install the dependencies
-
Get the API key from NFT.STORAGE and store that in end-to-end/packages/react-app/src/constants.js
-
Download the Neural Style Transfer Model from TensorFlow Hub and add the path to local directory in neural_style_transfer_api.py
-
Run the backend flask API
cd <path-to-NFT-Minting-dApp-with-Neural-Style-Transfer> python neural_style_transfer_api.py
-
Install Frontend modules using npm:
cd end-to-end yarn install
-
Start local hardhat devnet:
yarn chain
-
Open different terminal to deploy the smart contract:
yarn deploy
-
After smart contract deployment, run the UI:
yarn start
- Add ethers to the dummy account by clicking on "Grab Funds from the Faucet" on the UI.
- Add the content and the style image.
- Add NFT name and click on "Mint" button to add the generated image to IPFS and store IPFS URI on devnet. Note the token ID.
- Use the token ID to view the generated NFT in the "View an NFT" tab.