Hi, I have read your paper about the exclusion loss, does the exclusion loss can be written like this in pytorch:
(grad() is a function getting the H grad or V grad)
foreground_grad = torch.abs(grad(foreground))
background_grad = torch.abs(grad(background))
elemental_mul = foreground_grad.mul(background_grad)
loss = sum(elemental_mul)
Hi, I have read your paper about the exclusion loss, does the exclusion loss can be written like this in pytorch:
(grad() is a function getting the H grad or V grad) foreground_grad = torch.abs(grad(foreground)) background_grad = torch.abs(grad(background)) elemental_mul = foreground_grad.mul(background_grad) loss = sum(elemental_mul)
Looking forward to your reply.