LOFI: Whitepaper

Security and Privacy

We do not collect users' data without permission under any circumstances, now and in the future. In addition, it blocks the possibility of hacking and leakage through technically various layers.
The image input through the camera changed to an image that the user cannot identify via the dynamic vision processing method. On top of that, only landmarks such as eyes, nose, mouth, and hands are separately refined and converted into simple coordinates. This is called de-identification, and as a primary defense measure for personal information protection, even if the hacking attack of the user's PC web browser succeeds, the extracted data is not an image of the person's face, but only the coordinate information of the landmark.
In addition, since all data processing is handled within the user PC (Client-side) rather than our server-side, it is impossible to collect from the outside, so it is entirely free from hacking such as packet injection.
All data for broadcast transmission through the system that has undergone such a security process is limited to only the second layer, where the user's NFT and gesture are projected. Users fundamentally block all possibilities that their face or place may be leaked.