MCPcopy
hub / github.com/hzwer/ECCV2022-RIFE / forward

Method forward

model/oldmodel/IFNet_HDv2.py:64–86  ·  view source on GitHub ↗
(self, x, scale=1.0)

Source from the content-addressed store, hash-verified

62 self.block3 = IFBlock(10, scale=1, c=48)
63
64 def forward(self, x, scale=1.0):
65 if scale != 1.0:
66 x = F.interpolate(x, scale_factor=scale, mode="bilinear", align_corners=False)
67 flow0 = self.block0(x)
68 F1 = flow0
69 F1_large = F.interpolate(F1, scale_factor=2.0, mode="bilinear", align_corners=False) * 2.0
70 warped_img0 = warp(x[:, :3], F1_large[:, :2])
71 warped_img1 = warp(x[:, 3:], F1_large[:, 2:4])
72 flow1 = self.block1(torch.cat((warped_img0, warped_img1, F1_large), 1))
73 F2 = (flow0 + flow1)
74 F2_large = F.interpolate(F2, scale_factor=2.0, mode="bilinear", align_corners=False) * 2.0
75 warped_img0 = warp(x[:, :3], F2_large[:, :2])
76 warped_img1 = warp(x[:, 3:], F2_large[:, 2:4])
77 flow2 = self.block2(torch.cat((warped_img0, warped_img1, F2_large), 1))
78 F3 = (flow0 + flow1 + flow2)
79 F3_large = F.interpolate(F3, scale_factor=2.0, mode="bilinear", align_corners=False) * 2.0
80 warped_img0 = warp(x[:, :3], F3_large[:, :2])
81 warped_img1 = warp(x[:, 3:], F3_large[:, 2:4])
82 flow3 = self.block3(torch.cat((warped_img0, warped_img1, F3_large), 1))
83 F4 = (flow0 + flow1 + flow2 + flow3)
84 if scale != 1.0:
85 F4 = F.interpolate(F4, scale_factor=1 / scale, mode="bilinear", align_corners=False) / scale
86 return F4, [F1, F2, F3, F4]
87
88if __name__ == '__main__':
89 img0 = torch.zeros(3, 3, 256, 256).float().to(device)

Callers

nothing calls this directly

Calls 1

warpFunction · 0.90

Tested by

no test coverage detected