Send in your ideas. Deadline February 1, 2025
Grant
Theme fund: NGI0 PET
Period: 2019-10 — 2022-09
More projects like this
Services + Applications

Video chat privacy

Add privacy features to video chats

Making video calls can be very invasive to privacy: the camera does not only capture the face and posture of the person talking, but will in fact capture the entire environment in glorious high definition - from the books in your bookshelf to family members or laundry rack behind you. This information is of no interest to the other end, but with a camera you have little choice: once you slide open the camera cover, it takes everything within the field of view and broadcasts it to the other side. This project aims to use advanced AI technology to edit the video feed in real-time, and apply various privacy enhancements such as removal of backgrounds.

    Why does this actually matter to end users?

    One of the things people enjoy the most about the internet, is that it enables them to talk to others remotely almost without limit. Internet allows anyone to keep closely connected with friends and family, and help their kids solve a math problem while they are at work. People collaborate with their colleagues from the couch of their living room, the cafe where they enjoy lunch or on their cell phone on the bus to the gym. Businesses can easily service their customers where this is most convenient to them, without having to travel themselves. This is so convenient, that some businesses have already moved entirely online. Internet communication has become the nerve center of whole neighbourhoods, where people watch over the possessions of their neighbours while these are away for work or leisure.

    However, users have a hard time to understand how privacy is impacted if they use the wrong technology. Because internet works almost everywhere, the natural privacy protection of the walls of a house, a school or an office is gone. For example, when you make a video call, your webcam or phone camera captures a lot more than just you talking, for example the people around you, the books on your shelf or the street outside. A lot of this information can be used to uniquely identify you or to find your location, which you may not always be aware of. Because high definition cameras are embedded in more and more devices everywhere around us, we need more control over what these digital eyes actually record about us.

    Instead of only being able to switch your webcam on or off, this project will develop technology that lets you remove or anonymize the background while you are actually making a video call. This will give users more tools to protect their visual privacy and fight back against all sorts of sophisticated tracking schemes.

    Logo NLnet: abstract logo of four people seen from above Logo NGI Zero: letterlogo shaped like a tag

    This project was funded through the NGI0 PET Fund, a fund established by NLnet with financial support from the European Commission's Next Generation Internet programme, under the aegis of DG Communications Networks, Content and Technology under grant agreement No 825310.