Dice Loss

使用场景：解决样本不均衡的问题
优点：
缺点：
不稳定，容易出现梯度爆炸

实例化：dice_loss = loss_fns.SoftDiceLoss()
使用：loss1 = dice_loss(out, y)

class SoftDiceLoss(nn.Module):
    def __init__(self, weight=None, size_average=True, sigmoid=False):
        super(SoftDiceLoss, self).__init__()
        self.sigmoid = sigmoid

    def forward(self, logits, targets):
        num = targets.size(0)
        smooth = 1
        if self.sigmoid:
            logits = torch.sigmoid(logits)
        m1 = logits.view(num, -1)
        m2 = targets.view(num, -1)
        intersection = (m1 * m2)
        dice = 2. * (intersection.sum() + smooth) / ((m1 ** 2).sum() + (m1 ** 2).sum() + smooth)
        loss = 1. - dice.mean()
        return loss