/generate-to-adapt

PyTorch implementation of "Generate To Adapt: Aligning Domains using Generative Adversarial Networks"

Primary LanguagePython

Generate To Adapt

PyTorch implementation of "Generate To Adapt: Aligning Domains using Generative Adversarial Networks".

Datasets

Download the dataset from here and extract it. This folder contains the dataset in the same format as needed by the code.

Training

python train.py --data-root <path to dataset> --output-root <path to output results>

Current checkpoint and the best-performing model will be stored in the output directory provided.

Testing

python test.py --data-root <path to dataset>