KuroNet: Pre-Modern Japanese Kuzushiji Character Recognition with Deep Learning · HackerPapers discussions

KuroNet: Pre-Modern Japanese Kuzushiji Character Recognition with Deep Learning — mentioned by gyomu

Sun, 26 Apr 2026 10:23:43 GMT

If it makes you feel better, the vast majority of modern day Japanese speakers cannot read this either. It is cursive script, and only specialized academics/people with extensive training in calligraphy/etc. would know how to read it. Interestingly enough this is an area where machine learning has been extremely effective: https://arxiv.org/abs/1910.09433

Paper and discussion history. A mention is not an endorsement.

KuroNet: Pre-Modern Japanese Kuzushiji Character Recognition with Deep Learning — mentioned by macleginn

Wed, 03 Dec 2025 10:43:32 GMT

I became convinced of this after the release of KuroNet: https://arxiv.org/pdf/1910.09433 (High-quality OCR of Japanese manuscripts, which look almost impossible to read.)

Paper and discussion history. A mention is not an endorsement.

KuroNet: Pre-Modern Japanese Kuzushiji Character Recognition with Deep Learning — mentioned by czr

Sun, 17 Nov 2019 23:12:48 GMT

for those interested in the different ml methods attempted for the challenge, many of the top kagglers in this contest posted rundowns of their solution [0]. for those interested in kuronet [1] and the competition alex gave a presentation here (en, [2]) as did tarin (ja, [3]). for those interested in classical japanese literature, you can see the tools they've built here [4]. tarin is also very active on twitter [5]. [0] https://www.kaggle.com/c/kuzushiji-recognition/discussion [1] https://arxiv.org/pdf/1910.09433.pdf [2] https://youtu.be/e2-mVBvygm8?t=1428 [3]…

Paper and discussion history. A mention is not an endorsement.