some kanji cannot be recognized

[NCheng8835]

There are some kanji in 常用漢字表 that cannot be recognized.as follows:
曖傲璧哺刹嘲訃錮毀彙箋丼憬惧楷拉辣慄籠貪諧羞嗅瘍喩鬱摯緻恣踪

常用漢字表https://www.bunka.go.jp/kokugo_nihongo/sisaku/joho/joho/kijun/naikaku/pdf/joyokanjihyo_20101130.pdf

[ichisadashioko]

Sorry for not noticing the issue.
The current machine model for recognition was only able to recognize the characters from the ETL9B dataset. And it also doesn't have any katakana on it. I used ETL9B because it was the nicest and cleanest dataset out of the box. I do working on cleaning up all the datasets available to me in order to support more characters especially katakana.

The development should be scattered around these repositories.

• etlcdb - cleaning the datasets from a Japanese institute.
• hiragana-recognition - prototyping developing models and training pipeline with a small number of labels
• kanji-recognition - scale up the prototype with all the available datasets
• kanji-recognition-android - push the model to this repo for using in the android app

It is weird that github didn't send me any email notifying about new issues.