Deep learning - damage function (dice_loss)

'dice_loss' represents a loss function commonly used in image segmentation tasks, Dice Loss. The loss function is defined based on the Dice Coefficient (Dice Coefficient), which can be used to evaluate the similarity between the predicted segmented image and the real segmented image.

• In the binary classification problem, the calculation formula of Dice coefficient is:

It says = 2 * (|X ∩ Y|) / (|X| + |Y|)

Among them, X represents the positive pixel set in the predicted segmented image, Y represents the positive pixel set in the real segmented image, |X| and |Y| represent the size of the pixel set respectively, and |X ∩ Y| represents two pixel sets The intersection size of .

• The definition of Dice Loss is based on the reciprocal of the Dice coefficient, namely:

DiceLoss = 1 - Dice

Therefore, when the Dice coefficient is larger, the Dice Loss is smaller, indicating that the similarity between the predicted segmented image and the real segmented image is higher, and the performance of the model is better. Conversely, when the Dice coefficient is smaller, the Dice Loss is larger, indicating that the similarity between the predicted segmented image and the real segmented image is lower, and the performance of the model is worse.