Sources from here. It looks like many of these are not ML models, and more are just tools for, say, despekeling or timeseries analysis.
-
This states it does "change detection" in SAR imagery and has a torch.nn network and training loop.
-
This looks interesting and also contains an actual torch.nn network and training loop.
-
This does "patchwise" classification and contains the network and a training loop.
-
This looks like an actual detection model we can integrate with the paper here
-
Clarifi is a SAS tool to do classification. It takes a general model they have created and applies transfer learning. Looks closed source.
-
Labeled SAR imagery dataset of ten geophysical phenomena from Sentinel-1 wave mode might be a good dataset, but nothing about models.
-
This contains the detection of some SAR sat images, maybe we could adapt their model?
-
This is doing bounding boxes on HRSID but is all in notebooks and looks a little hard to generalize, under the hood it looks like its just applying YOLOv5 to HRSID.
Sources from here. It looks like many of these are not ML models, and more are just tools for, say, despekeling or timeseries analysis.
This states it does "change detection" in SAR imagery and has a torch.nn network and training loop.
This looks interesting and also contains an actual torch.nn network and training loop.
This does "patchwise" classification and contains the network and a training loop.
This looks like an actual detection model we can integrate with the paper here
Clarifi is a SAS tool to do classification. It takes a general model they have created and applies transfer learning. Looks closed source.
Labeled SAR imagery dataset of ten geophysical phenomena from Sentinel-1 wave mode might be a good dataset, but nothing about models.
This contains the detection of some SAR sat images, maybe we could adapt their model?
This is doing bounding boxes on HRSID but is all in notebooks and looks a little hard to generalize, under the hood it looks like its just applying YOLOv5 to HRSID.