Shalin Mehta

Replacing `viscy.unet.networks.unext2.UNext2Stem` with `StemDepthtoChannels` will break the compatibility between the constructed model and checkpoints for the virtual staining task.

It will be nice if this works out! We can also use the embedding viewer widget in napari-iohub to visualize the training examples and corresponding embeddings.

@edyoshikun I was able to draw a batch and construct the model! I now realize the model encodes temporal changes into a dense representation at each time point. The output...

@ziw-liu, please add tests using @aliddell's acquire-zarr example. Is there an example dataset with OME-NGFF 0.5 metadata from OME?

@alishbaimran @Soorya19Pradeep I am booked with multiple planning and writing projects through the end of the year. I think Soorya is situated to provide a solid review and testing of...

@alishbaimran it looks like you started a branch or merged changes in a way that affected files that are not related to this feature. You may need to do this:...

@ziw-liu, feature visualization vs. time didn't serve the purpose in early iterations. Did you also try it after we started training time-aware models? The temporal smoothness in the output layer...

To merge this branch and release candidate 0.3.0-rc1, we need to test the following: - [ ] Demos and notebooks that illustrate robust virtual staining: @ziw-liu @edyoshikun - [ ]...

> @mattersoflight I'm still working on https://github.com/mehta-lab/VisCy/pull/181 which will also introduce user interface changes. Should we do comprehensive release candidate testing after that? @ziw-liu I suggest merging #181 (CLI interface)...

@ziw-liu @Soorya19Pradeep @alishbaimran I agree with Ziwen - use UMAP for plotting/visualization and PCA for computing correlations with engineered features. We should do 3D UMAP renderings in napari, like in...