Created on 23rd July 2024
•
In an era where synthetic media and deepfake technology are becoming increasingly prevalent, distinguishing between real and manipulated visual content is a critical challenge. DeepCertify addresses this issue by providing a reliable and accessible tool for detecting deepfakes and issuing verifiable certificates. This service empowers users to make informed decisions about the credibility of multimedia content, thereby enhancing trust in digital media.
We decided to use Python for AI, and thought, why not run the entire backend with it? What could possibly go wrong, right? ðŸ«
We used Brownie for blockchain and FastAPI for the backend, but soon discovered that Brownie's requirements clashed with the backend's dependencies on multiple levels. 🫤 To fix this, we ended up creating two separate requirements.txt files to distribute the dependencies. ✅
Initially, we planned to mint the NFTs and show their URLs on OpenSea, so we wouldn't need to create an NFT marketplace. But we were too late to realize that OpenSea doesn't support zkEVM-Polygon. 😫 This meant we had to build our own NFT marketplace from scratch in Python just to display the certificates. 🚧
Making a good-looking certificate using Python was super hard, and we almost gave up on the design. 😓 Then we had a brilliant idea: create an HTML script with all the CSS and images, and using python take a screenshot to generate it as a certificate! 💡📜
Tracks Applied (1)
Polygon
Technologies used