Existing one-shot 4D head synthesis methods usually learn from monocular videos with the aid of 3DMM reconstruction, yet the latter is evenly challenging which restricts them from reasonable 4D head ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results