4DSHAPE will combine geometric methods and deep learning techniques to identify, develop, and perfect a natural framework where one can both embed and generate surfaces of the human body and faces independently of the way they are parameterized/discretized, including raw scans, in a way that captures and reproduces both the identity of the subject and the natural motions they can make.

The project will provide new tools, both theoretical and practical, not only for comparing 3D/4D surfaces but also for defining novel deformation processes for putting them in motion and/or changing their identity, irrespective of the way the input data may be parameterized. Including 4D data in our frameworks, we expect the results to open new horizons for several applications, including virtual reality, AR/VR content creation, gaming, and affective computing. See more