Inspiration
Out of our list of 32 ideas, and through a thorough process of convergent thinking techniques, we determined this idea fit the requirements the best, suited our interests, and had the highest value to effort ratio.
What it does
It is a prototype that demonstrates the idea of the fully implemented project. It detects whether an image of a person's face is real or fake, which protects the average consumer like you and me, from the malicious attempts at deepfaking... especially the ones where we might not even realize its fake!
It's an issue that only gets worse over time, so we should find a way that protects people, not just enterprises.
How we built it
We used a web extension to maximize the chances of the ability of installation and use. To get a good idea going, we used cradle-to-grave activity diagrams, user diagrams, and functional requirements. We used HTML, CSS, JavaScript, and JSON to develop it.
Challenges we ran into
Lack of API's False advertising of Deepfake API's Tried implementing our own AI only to get a virus and factory reseting a computer.
Accomplishments that we're proud of
Getting the prototype working on the demo site, as the idea is what is the most important aspect of this. Plus, our presentation is really well thought out and well designed.
What we learned
A lot of javascript, how to create a web extension, the use and availability of API's, the value of deepfake detection, and JSON.
What's next for Deepfake Defender
A working prototype (which could be done in a day's worth of work, which would be the time we typically would have for this hackathon had we not run into factory reseting a computer due to virus issues... plus more)
Log in or sign up for Devpost to join the conversation.