-
Notifications
You must be signed in to change notification settings - Fork 1.4k
Closed
Milestone
Description
Our multiple instance learning uses a list of patches (a bag of images) for training the model. Although we have a transform that splits the image into patches for our pipeline, this pipeline cannot be used for other use cases that do not rely on this transform, like this conversation: Project-MONAI/tutorials#896 (reply in thread). SplitDim is useful here but SplitDimd splits and stores into dictionary keys instead of list of dictionaries with the same key, which is required for this pipeline.
Metadata
Metadata
Assignees
Labels
No labels
Type
Projects
Status
💯 Complete