hardmaru / kanji2kanji

Reproduce domain transfer results in Deep Learning for Classical Japanese Literature

Kanji2Kanji

This repo contains instructions to reproduce the domain transfer results in the paper Deep Learning for Classical Japanese Literature.

Versions

Tested on TensorFlow 1.8.0, scipy 0.19.1, numpy 1.13.3, python 3.5

Instructions

First run python build_data.py to construct the Kanji dataset.

To train models, run python train.py to train all models, and run trained models on test set afterwards.

A notebook called demo.ipynb is also available to visualize the data and model predictions.

Code License

MIT

Data License

The stroke-based Kanji is derived from KanjiVG project.

The Kuzushiji Kanji data derived from the Kuzushiji-MNIST project.

Citation

If you find this work useful, we would appreciate a reference to our paper:

Deep Learning for Classical Japanese Literature. Tarin Clanuwat et al. arXiv:1812.01718

@online{clanuwat2018deep,
  author       = {Tarin Clanuwat and Mikel Bober-Irizar and Asanobu Kitamoto and Alex Lamb and Kazuaki Yamamoto and David Ha},
  title        = {Deep Learning for Classical Japanese Literature},
  date         = {2018-12-03},
  year         = {2018},
  eprintclass  = {cs.CV},
  eprinttype   = {arXiv},
  eprint       = {cs.CV/1812.01718},
}