Optimize DA

[jcohenadad]

https://github.com/sct-pipeline/contrast-agnostic-softseg-spinalcord/blob/1976cec36162f8dd1ce200cee13a3bf5493cbd9c/config_templates/meanGT_soft.json#L118

Ideas: artifacts, eg:

• RandomMotion RandomMotion RandomGhosting RandomGhosting RandomSpike RandomSpike

[naga-karthik]

We're sticking with the following data augmentation methods:

Training

Validation/Testing

I have noticed that adding more augmentations only increases the time per epoch with incremental improvements in performance.

EDIT: The transforms here are not up-to-date, please refer to the the next comment below.

[naga-karthik]

Here are the data augmentation transforms for the final model:

Training


Validation/Testing