FedDGH

implementation for **FedDGH: Federated Domain-Generalized Hashing via Dynamic Parameter Masking

Prerequisites:

• python == 3.12.2
• pytorch == 2.2.1
• torchvision == 0.17.1
• numpy, scipy, sklearn, PIL, argparse, flwr, timm, opencv-python

Dataset:

• Please manually download the datasets PACS, Office-Home, VLCS from the official websites.

Training:

cd FedDGH/source/main
./test.sh