customDiceForwardLoss.m File Reference

loss = customDiceForwardLoss(Y, T, dataDimension, useClasses) returns the Dice loss between the predictions Y and the training targets T. More...

Functions
function loss =	customDiceForwardLoss (Y, T, dataDimension, useClasses)
	loss = customDiceForwardLoss(Y, T, dataDimension, useClasses) returns the Dice loss between the predictions Y and the training targets T.

Detailed Description

loss = customDiceForwardLoss(Y, T, dataDimension, useClasses) returns the Dice loss between the predictions Y and the training targets T.

Function Documentation

customDiceForwardLoss()

function loss = customDiceForwardLoss	(		Y,
			T,
			dataDimension,
			useClasses )

loss = customDiceForwardLoss(Y, T, dataDimension, useClasses) returns the Dice loss between the predictions Y and the training targets T.

At the moment it is a test function to be used in [net, info] = trainnet(AugTrainDS, net, @customDiceForwardLoss, TrainingOptions); in startTraining dataDimensions

Parameters

Y	dlarray objects that correspond to the n network predictions (provided by trainnet)
T	dlarray objects that correspond to the n network targets (provided by trainnet)
dataDimension	value defining dimension of the data: 2, 2.5, 3
useClasses	indices of class ids to be used for calculation of loss, when empty, calculate for all classes

References N.