Attention is all you need

Inspiration

We want to provide an easy way to ensure the integrity of news and pieces of media through all the internet.

What it does

This application supervises the user's screen content searching for deep fake images and videos and alerts the user in the case of a detection.

How we built it

We've built this applications using python and pyQT to build the GUI and the alarm system, while we have also used an AI model to detect the deep fake instances.

Challenges we ran into

It has been hard to find a well-performing AI model, and since we didn't have any knowledge about QT, modifying the GUI to perform what we wanted specifically has been a challenge.

Accomplishments that we're proud of

We're mainly proud of the idea that we have come up with. It has the potential to be extended to other devices such as smartphones or browsers, and it can have a significant impact in the future over the integrity of information.

What we learned

We've learned a lot about state-of-the-art deep fake generation and detection models, as well as QT programming.

What's next for Attention is all you need

As I mentioned above, extending this application to other devices or browsers might multiply its impact in the future with little effort. It would also be interesting to extend the application to detect deep fake audio in the future.

Github repo

You can find the code of the project in this repo: https://github.com/Bimo99B9/Big-Hackobowl

Built With

• python
• pytorch
• qt