DigiFace

The product we aim for is a program that can overlay certain images on the face that a camera sees and makes the user see that overlay on the face in real time. The idea is totally like SnapChat Lenses, but we are leaving the feature of being a 3D AR overlay optional depending on the time remaining, and, of course in the discussion, the TA can guide us to what we should make optional and what we must implement. What we think we must implement up till now is overlaying a 2D image on a single face that resizes different aspects of it depending on the changes of the eyes and mouth sizes, and that works in different lighting conditions ranging from outdoors in the sunlight to indoors in a normal room lighting for different face orientations and using the quality of a conventional front facing camera. The optional features that could be added are being able to trigger certain events depending on certain motions that the user does like shaking his head, being a 3D model that changes its orientation depending on the orientation of the face and perfectly wraps the face, allowing multiple faces, recognizing the face and thus recommending certain overlays, recognizing gender, and, finally, adding a feature like SnapChat World Lens that detects flat surfaces and puts different 3D objects on it as if it is resting on the surface.