Implementation of LVSM, SOTA Large View Synthesis with Minimal 3d Inductive Bias, from Adobe Research
