You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If you mean target coordinates (x,y) then it is quite simple. Just take the starting position of that pixel and add the flow vector of that pixel to get the target coordinates:
#assuming flow has the shape BxCxHxW
yy,xx = torch.meshgrid(torch.arange(flow.size(-2)), torch.arange(flow.size(-1)), indexing="ij")
spatial_pos = torch.stack([xx,yy], dim=0)
target_pos = spatial_pos.unsqueeze(0)+flow
Thanks for your nice work!
I'm curious about how to translate flow to coordinate. (u,v -> x, y)
The text was updated successfully, but these errors were encountered: