def test_sigmoid_focal_loss():
input_good = torch.Tensor([10, -10, 10]).float()
input_bad = torch.Tensor([-1, 2, 0]).float()
target = torch.Tensor([1, 0, 1])
loss_good = F.sigmoid_focal_loss(input_good, target)
loss_bad = F.sigmoid_focal_loss(input_bad, target)
assert loss_good < loss_bad
def forward(self, input: torch.Tensor, target: torch.Tensor):
if self.ignore_index is not None:
mask = target != self.ignore_index
target = target[mask]
input = input[mask]
if not len(target):
return torch.tensor(0.).to(input.device)
focal_loss = 0
num_classes = input.size(1)
for cls in range(num_classes):
cls_label_target = (target == cls).long()
cls_label_input = input[:, cls]
focal_loss += sigmoid_focal_loss(cls_label_input,
cls_label_target,
gamma=self.gamma,
alpha=None)
# Second term
y = F.log_softmax(input, dim=1).exp()
target_one_hot = F.one_hot(target, input.size(1)).float()
# +1 to make loss be [0;2], instead [-1;1]
kappa_loss = 1 + quad_kappa_loss_v2(
y, target_one_hot, y_pow=self.y_pow, eps=self.eps)
return kappa_loss + self.log_scale * focal_loss
def forward(self, label_input, label_target):
num_classes = label_input.size(1)
loss = 0
# Filter anchors with -1 label from loss computation
if self.ignore_index is not None:
not_ignored = label_target != self.ignore_index
label_input = label_input[not_ignored]
label_target = label_target[not_ignored]
for cls in range(num_classes):
cls_label_target = (label_target == cls).long()
cls_label_input = label_input[:, cls]
loss += sigmoid_focal_loss(cls_label_input, cls_label_target, gamma=self.gamma, alpha=self.alpha)
return loss