Generalizable Neural Performer

Learning Robust Radiance Fields for Human Novel View Synthesis