Hi!
Im experimenting with fully automated workflow that takes a directory of images and optional head mesh to create metahuman with texture that looks correct in different lighting. Basically projecting the images correctly and also not bake in light. Its markerless even though images here has them. The usecase might be to have AI generated faces and meshes. The ones here are from triplegangers. As seen it has difficulties with eyelids, nose and ears, but that might be easier to correct manually than doing the full face. Im wondering if there is something like this already?
