search

awei669 / VQ-Font

[ICCV 2023] Few shot font generation via transferring similarity guided global and quantization local styles
https://arxiv.org/abs/2309.00827
128 stars 6 forks source link

About datasets #1

Closed striver123456 closed 1 year ago

striver123456 commented 1 year ago

Great Work!!! How did you create the dataset?

awei669 commented 1 year ago

Thank you for your interest in our work！

  1. You should collect different ttf files, determine the source font, and divide the training set and test set.
  2. Determine the generated character set (it also needs to be divided into training character set ‘train_unis’ and test character set ‘val_unis’), use font2image.py to generate images, see readme for organizational form.
  3. Run script ./build_trainset.sh.

Hope can help you～

awei669 commented 1 year ago

If you have any further questions, re-open the issue.