- Implementation of MFDE (Mesoscale eddy detection based on multi-feature fusion network)
-
SSH data: ADT Dataset )
-
SST: [SST Dataset ]
-
Our data(SSH,SST and Contour data): [Google Drive]
-
Best_model:[Google Drive]
- Download the data, setup data-paths in the datasets
- Use the training scripts[train.py] for paired training of MFED
The Overall Flowchart of the Proposed Architecture |
---|
Mesoscale eddy semantic segmentation task |
---|
Mesoscale eddy contours detection task |
---|