Resources

  • Implementation of MFDE (Mesoscale eddy detection based on multi-feature fusion network)

Datasets

  1. SSH data: ADT Dataset )

  2. SST: [SST Dataset ]

  3. Our data(SSH,SST and Contour data): [Google Drive]

  4. Best_model:[Google Drive]

Usage

  • 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
det-enh
Mesoscale eddy semantic segmentation task
det-enh
Mesoscale eddy contours detection task
det-enh

Acknowledgements