MCPcopy Create free account
hub / github.com/VisionXLab/OF-Diff / delta_border

Method delta_border

ldm/models/diffusion/ddpm.py:692–704  ·  view source on GitHub ↗

:param h: height :param w: width :return: normalized distance to image border, wtith min distance = 0 at border and max dist = 0.5 at image center

(self, h, w)

Source from the content-addressed store, hash-verified

690 return arr
691
692 def delta_border(self, h, w):
693 """
694 :param h: height
695 :param w: width
696 :return: normalized distance to image border,
697 wtith min distance = 0 at border and max dist = 0.5 at image center
698 """
699 lower_right_corner = torch.tensor([h - 1, w - 1]).view(1, 1, 2)
700 arr = self.meshgrid(h, w) / lower_right_corner
701 dist_left_up = torch.min(arr, dim=-1, keepdims=True)[0]
702 dist_right_down = torch.min(1 - arr, dim=-1, keepdims=True)[0]
703 edge_dist = torch.min(torch.cat([dist_left_up, dist_right_down], dim=-1), dim=-1)[0]
704 return edge_dist
705
706 def get_weighting(self, h, w, Ly, Lx, device):
707 weighting = self.delta_border(h, w)

Callers 1

get_weightingMethod · 0.95

Calls 1

meshgridMethod · 0.95

Tested by

no test coverage detected