You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Now we have switched to using Segformer models by default, which are a different model architecture to Unets, it is potentially confusing that we use the name SDS_unet_classifier.ipynb
I know I have requested name changes in the past (sorry!) but to standardize it, and link it recursively back to the Doodleverse, I propose we rename it SDS_zoo_classifier.ipynb because the models are hosted in Segmentation Zoo and were made using Gym. I just think it will clarify things for some folks.
This is obviously not urgent and can happen when we implement the new shoreline extraction logic for the models in #168
thanks!
The text was updated successfully, but these errors were encountered:
A related request would be to update the README, so after 5. Use the SDS_coastsat_classifier to Download Imagery, perhaps put the following?
6. Use the SDS_zoo_classifier to use a deep learning based segmentation model for shoreline extraction. You must have already downloaded imagery using SDS_coastsat_classifier . Warning: this program is still beta and subject to many and frequent changes. We will announce when the first stable version of this program is available, so in the meantime, use at your own risk!
(or something like that! Not a bad idea to put that in the top of the notebook too! Just for now ....)
Now we have switched to using Segformer models by default, which are a different model architecture to Unets, it is potentially confusing that we use the name
SDS_unet_classifier.ipynb
I know I have requested name changes in the past (sorry!) but to standardize it, and link it recursively back to the Doodleverse, I propose we rename it
SDS_zoo_classifier.ipynb
because the models are hosted in Segmentation Zoo and were made using Gym. I just think it will clarify things for some folks.This is obviously not urgent and can happen when we implement the new shoreline extraction logic for the models in #168
thanks!
The text was updated successfully, but these errors were encountered: