Abstract: Recent advances in implicit neural representations make it possible to generate free-viewpoint videos of the human from sparse view images. To avoid the expensive training for each person, ...