Validaiton result is not good,how to improve it?

#1

Hi,i want to use Unet segment target in different dataset.In Unet,using GroupNormalization,LeakRelu,maxpooling and conv,traing opt is Adam,loss is mutildice,accu is mean dice.
first i segment five target,the train and validation performence are both good,look first line:
second i segment three target,the train performence is good,but validation is not well,look second line:

image
image1009×759 113 KB

can anyone know how to solve it?

thank you!!!
chenjunqiang

#2

you need to add the regularization i.e L1, L2 . If the results were not improved you can increase the dataset by using data augmentation

#3

i have add data augmentation with zoom, pan, rotate, flip operation，i will try regularization L2 to Adam opt.

#4

you also need to add the dropout layer as your model is clearly over fitted

#5

ok，i also add dropout layer after conv operation.

#6

i add dropout,but the result is also not good.

image
image899×344 51.1 KB