Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation
This addresses the limitation of low-dimensional subspace methods in face geometry reconstruction, offering a more expressive solution for applications in computer vision and graphics.
The paper tackles the problem of reconstructing facial geometry from a single image by proposing an image-to-image translation network that maps to depth and correspondence maps, enabling high-quality reconstructions for diverse faces under extreme expressions, with qualitative and quantitative analyses showing accuracy and robustness.
It has been recently shown that neural networks can recover the geometric structure of a face from a single given image. A common denominator of most existing face geometry reconstruction methods is the restriction of the solution space to some low-dimensional subspace. While such a model significantly simplifies the reconstruction problem, it is inherently limited in its expressiveness. As an alternative, we propose an Image-to-Image translation network that jointly maps the input image to a depth image and a facial correspondence map. This explicit pixel-based mapping can then be utilized to provide high quality reconstructions of diverse faces under extreme expressions, using a purely geometric refinement process. In the spirit of recent approaches, the network is trained only with synthetic data, and is then evaluated on in-the-wild facial images. Both qualitative and quantitative analyses demonstrate the accuracy and the robustness of our approach.