| 6 | |
| 7 | |
| 8 | def warp(tenInput, tenFlow, device): |
| 9 | backwarp_tenGrid = {} |
| 10 | k = (str(tenFlow.device), str(tenFlow.size())) |
| 11 | if k not in backwarp_tenGrid: |
| 12 | tenHorizontal = torch.linspace(-1.0, 1.0, tenFlow.shape[3], device=device).view( |
| 13 | 1, 1, 1, tenFlow.shape[3]).expand(tenFlow.shape[0], -1, tenFlow.shape[2], -1) |
| 14 | tenVertical = torch.linspace(-1.0, 1.0, tenFlow.shape[2], device=device).view( |
| 15 | 1, 1, tenFlow.shape[2], 1).expand(tenFlow.shape[0], -1, -1, tenFlow.shape[3]) |
| 16 | backwarp_tenGrid[k] = torch.cat( |
| 17 | [tenHorizontal, tenVertical], 1).to(device) |
| 18 | |
| 19 | tenFlow = torch.cat([tenFlow[:, 0:1, :, :] / ((tenInput.shape[3] - 1.0) / 2.0), |
| 20 | tenFlow[:, 1:2, :, :] / ((tenInput.shape[2] - 1.0) / 2.0)], 1) |
| 21 | |
| 22 | g = (backwarp_tenGrid[k] + tenFlow).permute(0, 2, 3, 1) |
| 23 | return torch.nn.functional.grid_sample(input=tenInput, grid=g, mode='bilinear', padding_mode='border', align_corners=True) |
| 24 | |
| 25 | |
| 26 | def conv(in_planes, out_planes, kernel_size=3, stride=1, padding=1, dilation=1): |