StyleRig: Rigging StyleGAN for 3D Control over Portrait Images
This provides a solution for applications requiring realistic and controllable face images, such as animation or virtual reality, though it is incremental by combining existing models.
The paper tackles the problem of lacking 3D semantic control over photorealistic portrait images generated by StyleGAN, and presents a method that enables explicit control over face pose, expressions, and illumination while maintaining photorealism.
StyleGAN generates photorealistic portrait images of faces with eyes, teeth, hair and context (neck, shoulders, background), but lacks a rig-like control over semantic face parameters that are interpretable in 3D, such as face pose, expressions, and scene illumination. Three-dimensional morphable face models (3DMMs) on the other hand offer control over the semantic parameters, but lack photorealism when rendered and only model the face interior, not other parts of a portrait image (hair, mouth interior, background). We present the first method to provide a face rig-like control over a pretrained and fixed StyleGAN via a 3DMM. A new rigging network, RigNet is trained between the 3DMM's semantic parameters and StyleGAN's input. The network is trained in a self-supervised manner, without the need for manual annotations. At test time, our method generates portrait images with the photorealism of StyleGAN and provides explicit control over the 3D semantic parameters of the face.