Skip to content

Conversation

@jacob-rosenthal
Copy link
Collaborator

Add U-net, plus tests.
This is the model that they used in the PESO paper

Copy link
Contributor

@ryanccarelli ryanccarelli left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's add to our contributing that models must reproduce baselines from paper before merge

* - :class:`~pathml.ml.unet.UNet`
- A standard general-purpose model designed for segmentation in biomedical images.
Architecture consists of an 4 encoder blocks followed by 4 decoder blocks.
Skip connections propagate information from each layer of the encoder to the corresponding layer in
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

layer of same dimension

- A model for simultaneous nucleus segmentation and classification in H&E images.
Architecture consists of a single encoder with three separate decoder branches: one to perform binary
classification of nuclear pixels (NP), one to compute horizontal and vertical nucleus maps (HV), and one which
is used in the classification setting to perform classification of nuclear pixels (NC).
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

verbose

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants