Synthesis from single view naturally suffers from bas-relief ambiguity making it much more challenging, but our prior model enables plausible results. The bottom left image was taken from the Rodin website. The other images were taken with Pixel smartphones.

We reconstruct examples from LP3D (Trevithick et al., 2023). Note that our prior model has never seen any stylization or strong face paint.

We show one-shot results on Multiface subjects.