🔥 Disentangling Writer and Character Styles for Handwriting Generation

📢 Introduction

The proposed style-disentangled Transformer (SDT) generates online handwritings with conditional content and style. Existing RNN-based methods mainly focus on capturing a person’s overall writing style, neglecting subtle style inconsistencies between characters written by the same person. In light of this, SDT disentangles the writer-wise and character-wise style representations from individual handwriting samples for enhancing imitation performance.
We extend SDT and introduce an offline-to-offline framework for improving the generation quality of offline Chinese handwritings.

📺 Handwriting generation results

Online Chinese handwriting generation
Applications to various scripts
Extension on offline Chinese handwriting generation

🔨 Requirements

python 3.8
pytorch >=1.8
easydict 1.9
einops 0.4.1

📂 Folder Structure

SDT/
│
├── train.py - main script to start training
├── test.py - generate characters via trained model
├── evaluate.py - evaluation of generated samples
│
├── configs/*.yml - holds configuration for training
├── parse_config.py - class to handle config file
│
├── data_loader/ - anything about data loading goes here
│   └── loader.py
│
├── model_zoo/ - pre-trained content encoder model
│
├── data/ - default directory for storing experimental datasets
│
├── model/ - networks, models and losses
│   ├── encoder.py
│   ├── gmm.py
│   ├── loss.py
│   ├── model.py
│   └── transformer.py
│
├── saved/
│   ├── models/ - trained models are saved here
│   ├── tborad/ - tensorboard visualization
│   └── samples/ - visualization samples in the training process
│
├── trainer/ - trainers
│   └── trainer.py
│  
└── utils/ - small utility functions
    ├── util.py
    └── logger.py - set log dir for tensorboard and logging output

💿 Datasets

We provide Chinese, Japanese and English datasets in data. Please download these datasets, uzip them and move the extracted files to /data.

🍔 Pre-trained model

We provide the pre-trained content encoder model in model_zoo. Please download and put it to the /model_zoo.
We provide the well-trained SDT model in saved_weights so that users can get rid of retraining one and play it right away.

🚀 Training & Test

Training

To train the SDT on the Chinese dataset, run this command:

python train.py --cfg configs/CHINESE_CASIA.yml --log Chinese_log

To train the SDT on the Japanese dataset, run this command:

python train.py --cfg configs/Japanese_TUATHANDS.yml --log Japanese_log

To train the SDT on the English dataset, run this command:

python train.py --cfg configs/English_CASIA.yml --log English_log

Qualitative Test

To generate Chinese handwritings with our SDT, run this command:

python test.py --pretrained_model checkpoint_path --dir Generated/Chinese

To generate Japanese handwritings with our SDT, run this command:

python test.py --pretrained_model checkpoint_path --dir Generated/Japanese

To generate English handwritings with our SDT, run this command:

python test.py --pretrained_model checkpoint_path --dir Generated/English

Quantitative Evaluation

To evaluate the generated handwritings, you need to set data_path to the path of the generated handwritings (e.g., Generated/Chinese), and run this command:

python evaluate.py --data_path Generated/Chinese

❤️ Citation

If you find our work inspiring or use our codebase in your research, please cite our work:

@article{dai2023disentangling,
  title={Disentangling Writer and Character Styles for Handwriting Generation},
  author={Dai, Gang and Zhang, Yifan and Wang, Qingfeng and Du, Qing and Yu, Zhuliang and Liu, Zhuoman and Huang, Shuangping},
  journal={arXiv preprint arXiv:2303.14736},
  year={2023}
}

robotPin / SDT