Closed posobin closed 7 years ago
In laplacian generation procedure (getlaplacian1), vector np.mean(win_i, axis=0) has size c, not win_size * 2 + 1 (even though in deep photo case they are both equal to 3).
getlaplacian1
np.mean(win_i, axis=0)
win_size * 2 + 1
In laplacian generation procedure (
getlaplacian1
), vectornp.mean(win_i, axis=0)
has size c, notwin_size * 2 + 1
(even though in deep photo case they are both equal to 3).