Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: ja
|
| 3 |
+
tags:
|
| 4 |
+
- image-to-text
|
| 5 |
+
license: apache-2.0
|
| 6 |
+
datasets:
|
| 7 |
+
- manga109s
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# Manga OCR
|
| 11 |
+
|
| 12 |
+
Optical character recognition for Japanese text, with the main focus being Japanese manga.
|
| 13 |
+
|
| 14 |
+
It uses [Vision Encoder Decoder](https://huggingface.co/docs/transformers/model_doc/visionencoderdecoder) framework.
|
| 15 |
+
|
| 16 |
+
Manga OCR can be used as a general purpose printed Japanese OCR, but its main goal was to provide a high quality
|
| 17 |
+
text recognition, robust against various scenarios specific to manga:
|
| 18 |
+
- both vertical and horizontal text
|
| 19 |
+
- text with furigana
|
| 20 |
+
- text overlaid on images
|
| 21 |
+
- wide variety of fonts and font styles
|
| 22 |
+
- low quality images
|
| 23 |
+
|
| 24 |
+
Code for inference is available [here](https://github.com/kha-white/manga_ocr).
|
| 25 |
+
|
| 26 |
+
Code for training will be released soon.
|