synthtiger 1.0.0

Synthetic text image generator for OCR model

🐯 SynthTIGER: Synthetic Text Image Generator

Paper | Documentation | Datasets

SynthTIGER is synthetic text image generator for OCR model.

Contents

• Documentation
• Installation
• Usage
• Advanced Usage
• Datasets
• Citation
• License

Documentation

The documentation is available at https://clovaai.github.io/synthtiger/.

You can check API reference in this documentation.

Installation

SynthTIGER requires python>=3.6 and libraqm. If you want install dependencies, see dependencies.

To install SynthTIGER from PyPI:

$ pip install synthtiger

Usage

# Set environment variable (for macOS)
$ export OBJC_DISABLE_INITIALIZE_FORK_SAFETY=YES

usage: synthtiger [-h] [-o PATH] [-c INTEGER] [-w INTEGER] [-v] SCRIPT NAME [CONFIG]

positional arguments:
  SCRIPT                Script file path.
  NAME                  Template class name.
  CONFIG                Config file path.

optional arguments:
  -h, --help            show this help message and exit
  -o PATH, --output PATH
                        Directory path to save data.
  -c INTEGER, --count INTEGER
                        Number of data. [default: 100]
  -w INTEGER, --worker INTEGER
                        Number of workers. If 0, It generates data in the main process. [default: 0]
  -v, --verbose         Print error messages while generating data.

Examples

SynthTIGER text images

# horizontal
synthtiger -o results -w 4 -v examples/synthtiger/template.py SynthTiger examples/synthtiger/config_horizontal.yaml

# vertical
synthtiger -o results -w 4 -v examples/synthtiger/template.py SynthTiger examples/synthtiger/config_vertical.yaml

Multiline text images

synthtiger -o results -w 4 -v examples/multiline/template.py Multiline examples/multiline/config.yaml

Advanced Usage

Non-Latin language data generation

1. Prepare corpus

   txt format, line by line. (example)

2. Prepare fonts

   See font customization for more details.

3. Edit corpus path and font path in config file

4. Run gen.py

Font customization

1. Prepare fonts

   ttf/otf format. (example)

2. Extract renderable charsets

   python tools/extract_font_charset.py -w 4 fonts/
   

   This script extracts renderable charsets for all font files. (example)

   Text files are generated in the input path with the same names as the fonts.

3. Edit font path in config file

4. Run gen.py

Colormap customization

1. Prepare images

   jpg/jpeg/png/bmp format.

2. Create colormaps

   python tools/create_colormap.py --max_k 3 -w 4 images/ colormap.txt
   

   This script creates colormaps for all image files. (example)

3. Edit colormap path in config file

4. Run gen.py

Template customization

You can implement custom templates by inheriting the base template.

from synthtiger import templates


class MyTemplate(templates.Template):
    def __init__(self, config=None):
        # initialize template.

    def generate(self):
        # generate data.

    def init_save(self, root):
        # initialize something before save.

    def save(self, root, data, idx):
        # save data to specific path.

    def end_save(self, root):
        # finalize something after save.

Datasets

SynthTIGER is available for download at google drive.

Dataset was split into several smaller files. Please download all files and run following command.

# for Linux, macOS
cat synthtiger_v1.0.zip.* > synthtiger_v1.0.zip

# for Windows
copy /b synthtiger_v1.0.zip.* synthtiger_v1.0.zip

synthtiger_v1.0.zip (36G) (md5: 5b5365f4fe15de24e403a9256079be70)

• Original paper version.

synthtiger_v1.1.zip (38G) (md5: b2757a7e2b5040b14ed64c473533b592)

• Used MJ/ST lexicon instead of MJ/ST label.
• Fixed a bug that applies transformation twice on curved text.
• Fixed a bug that incorrectly converts grayscale to RGB.

Version	IIIT5k	SVT	IC03	IC13	IC15	SVTP	CUTE80	Total
1.0	93.2	87.3	90.5	92.9	72.1	77.7	80.6	85.9
1.1	93.4	87.6	91.4	93.2	73.9	77.8	80.6	86.6

Structure

The structure of the dataset is as follows. The dataset contains 10M images.

gt.txt
images/
    0/
        0.jpg
        1.jpg
        ...
        9998.jpg
        9999.jpg
    1/
    ...
    998/
    999/

The format of gt.txt is as follows. Image path and label are separated by tab. (<image_path>\t<label>)

images/0/0.jpg	10
images/0/1.jpg	date:
...
images/999/9999998.jpg	STUFFIER
images/999/9999999.jpg	Re:

Citation

@inproceedings{yim2021synthtiger,
  title={SynthTIGER: Synthetic Text Image GEneratoR Towards Better Text Recognition Models},
  author={Yim, Moonbin and Kim, Yoonsik and Cho, Han-Cheol and Park, Sungrae},
  booktitle={International Conference on Document Analysis and Recognition},
  pages={109--124},
  year={2021},
  organization={Springer}
}

License

SynthTIGER
Copyright (c) 2021-present NAVER Corp.

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.

The following directories and their subdirectories are licensed the same as their origins. Please refer to NOTICE

docs/
resources/font/