popup close
DuckDuckGoose
Team info
Joris Mollinga
Netherlands (Nederland)
Bachelor
Mark Evenblij
Netherlands (Nederland)
Bachelor
Parya Lotfi
Netherlands (Nederland)
Bachelor
Category
Clusters
We are looking for new team members with the following skills

Experience with working with neural networks, knowledge of generating data and a passion to debunk deepfake threats. Preferably a data scientist, but if you are a skilled in Python, PyTorch or OpenCV we are very interested as well!

Contact us

DuckDuckGoose

The challenge

Would you like to play in a Star Wars movie, that's possible with a push of a button. Would I like to blackmail you by framing you in a child porn video? That as well is possible with a push of a button. Well, don't worry. At the moment you can still debunk these deepfake video's with your own eyes. But if this technology will develop. How will the media fight fake news, if they don't know what's true. How will the state assure justice, if they don't know if the proof is actually legitimate?

The solution

The DeepDetector is a neural network that detects the deepfakes and breaks the artifacts down. We from DuckDuckGoose are building this tool to empower analysists in their decision making and the justification to judge and public. With a subscription based model we want to make our product available to national institutions who can run their data anonymously through our software. By debunking fake news with full transparancy we want to ensure the digital world stays one we can still believe in.

Photo's
Videos
This project is being coached by

Contact

tudelftcontest@soapbox.nl

Tel: +31 (0)6 38 09 08 92 (WhatsApp)

TU Delft

Postbus 5

2600 AA Delft

Follow us