FW-GAN is a frequency-aware, one-shot handwriting synthesis framework designed to produce realistic and writer-consistent handwritten text from a single reference public at
Expert Systems with Applications
Training code is released on
GitHub.
1conda create --name fwgan python=3.10
2conda activate fwgan
3pip3 install torch torchvision --index-url https://download.pytorch.org/whl/cu126
4git clone https://github.com/DAIR-Group/FW-GAN.git && cd FW-GAN
5pip install -r requirements.txt
We provide our pretrained model weights and datasets here. For training, please download the files the files
train.hdf5 and
test.hdf5 and place them into the
data folder. For quick evaluation, you can also download the pretrained model
FW-GAN.pth and place it under
/data/weights/FW-GAN.pth on the code released on
GitHub.
This work is partially based on the code released for
HiGAN1@article{TONGDANGKHOA2026130175,
2title = {FW-GAN: Frequency-driven handwriting synthesis with wave-modulated MLP generator},
3journal = {Expert Systems with Applications},
4volume = {299},
5pages = {130175},
6year = {2026},
7issn = {0957-4174},
8doi = {https://doi.org/10.1016/j.eswa.2025.130175},
9url = {https://www.sciencedirect.com/science/article/pii/S095741742503790X},
10author = {Huynh {Tong Dang Khoa} and Dang {Hoai Nam} and Vo {Nguyen Le Duy}},
11keywords = {Handwritten text synthesis, Wavelet transform, One-shot learning, Vietnamese handwriting, Synthetic data, Generative adversarial networks},
12}
