Cross-Language Transfer-Learning Approach via a Pretrained Preact ResNet-18 Architecture for Improving Kanji Recognition Accuracy and Enhancing a Number of Recognizable Kanji

Many people admire the Japanese language and culture, but mastering the language’s writing system, particularly handwritten kanji, presents a significant challenge. Furthermore, translating historical manuscripts containing archaic or rare kanji requires specialized expertise. To address this, we de...

Full description

Saved in:
Bibliographic Details
Main Authors: Vasyl Rusyn, Andrii Boichuk, Lesia Mochurad
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/9/4894
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850277604783816704
author Vasyl Rusyn
Andrii Boichuk
Lesia Mochurad
author_facet Vasyl Rusyn
Andrii Boichuk
Lesia Mochurad
author_sort Vasyl Rusyn
collection DOAJ
description Many people admire the Japanese language and culture, but mastering the language’s writing system, particularly handwritten kanji, presents a significant challenge. Furthermore, translating historical manuscripts containing archaic or rare kanji requires specialized expertise. To address this, we designed a new model for handwritten kanji recognition based on the concept of cross-language transfer learning using a Preact ResNet-18 architecture. The model was pretrained in a Chinese dataset and subsequently fine-tuned in a Japanese dataset. We also adapted and evaluated two fine-tuning strategies: unfreezing only the last layer and unfreezing all the layers during fine-tuning. During the implementation of our training algorithms, we trained a model with the CASIA-HWDB dataset with handwritten Chinese characters and used its weights to initialize models that were fine-tuned with a Kuzushiji-Kanji dataset that consists of Japanese handwritten kanji. We investigated the effectiveness of the developed model when solving a multiclass classification task for three subsets with the one hundred fifty, two hundred, and three hundred most-sampled classes and showed an improvement in the recognition accuracy and an enhancement in a number of recognizable kanji with the proposed model compared to those of the existing methods. Our best model achieved 97.94% accuracy for 150 kanji, exceeding the previous SOTA result by 1.51%, while our best model for 300 kanji achieved 97.62% accuracy (exceeding the 150-kanji SOTA accuracy by 1.19% while doubling the class count). This confirms the effectiveness of our proposed model and establishes new benchmarks in handwritten kanji recognition, both in terms of accuracy and the number of recognizable kanji.
format Article
id doaj-art-3cae478c1ff147e496f9c359cc4b5029
institution OA Journals
issn 2076-3417
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-3cae478c1ff147e496f9c359cc4b50292025-08-20T01:49:50ZengMDPI AGApplied Sciences2076-34172025-04-01159489410.3390/app15094894Cross-Language Transfer-Learning Approach via a Pretrained Preact ResNet-18 Architecture for Improving Kanji Recognition Accuracy and Enhancing a Number of Recognizable KanjiVasyl Rusyn0Andrii Boichuk1Lesia Mochurad2Department of Artificial Intelligence, Lviv Polytechnic National University, 79013 Lviv, UkraineDepartment of Artificial Intelligence, Lviv Polytechnic National University, 79013 Lviv, UkraineDepartment of Artificial Intelligence, Lviv Polytechnic National University, 79013 Lviv, UkraineMany people admire the Japanese language and culture, but mastering the language’s writing system, particularly handwritten kanji, presents a significant challenge. Furthermore, translating historical manuscripts containing archaic or rare kanji requires specialized expertise. To address this, we designed a new model for handwritten kanji recognition based on the concept of cross-language transfer learning using a Preact ResNet-18 architecture. The model was pretrained in a Chinese dataset and subsequently fine-tuned in a Japanese dataset. We also adapted and evaluated two fine-tuning strategies: unfreezing only the last layer and unfreezing all the layers during fine-tuning. During the implementation of our training algorithms, we trained a model with the CASIA-HWDB dataset with handwritten Chinese characters and used its weights to initialize models that were fine-tuned with a Kuzushiji-Kanji dataset that consists of Japanese handwritten kanji. We investigated the effectiveness of the developed model when solving a multiclass classification task for three subsets with the one hundred fifty, two hundred, and three hundred most-sampled classes and showed an improvement in the recognition accuracy and an enhancement in a number of recognizable kanji with the proposed model compared to those of the existing methods. Our best model achieved 97.94% accuracy for 150 kanji, exceeding the previous SOTA result by 1.51%, while our best model for 300 kanji achieved 97.62% accuracy (exceeding the 150-kanji SOTA accuracy by 1.19% while doubling the class count). This confirms the effectiveness of our proposed model and establishes new benchmarks in handwritten kanji recognition, both in terms of accuracy and the number of recognizable kanji.https://www.mdpi.com/2076-3417/15/9/4894deep learningmulticlass classificationlanguage recognitionResNetPreact ResNet-18computer vision
spellingShingle Vasyl Rusyn
Andrii Boichuk
Lesia Mochurad
Cross-Language Transfer-Learning Approach via a Pretrained Preact ResNet-18 Architecture for Improving Kanji Recognition Accuracy and Enhancing a Number of Recognizable Kanji
Applied Sciences
deep learning
multiclass classification
language recognition
ResNet
Preact ResNet-18
computer vision
title Cross-Language Transfer-Learning Approach via a Pretrained Preact ResNet-18 Architecture for Improving Kanji Recognition Accuracy and Enhancing a Number of Recognizable Kanji
title_full Cross-Language Transfer-Learning Approach via a Pretrained Preact ResNet-18 Architecture for Improving Kanji Recognition Accuracy and Enhancing a Number of Recognizable Kanji
title_fullStr Cross-Language Transfer-Learning Approach via a Pretrained Preact ResNet-18 Architecture for Improving Kanji Recognition Accuracy and Enhancing a Number of Recognizable Kanji
title_full_unstemmed Cross-Language Transfer-Learning Approach via a Pretrained Preact ResNet-18 Architecture for Improving Kanji Recognition Accuracy and Enhancing a Number of Recognizable Kanji
title_short Cross-Language Transfer-Learning Approach via a Pretrained Preact ResNet-18 Architecture for Improving Kanji Recognition Accuracy and Enhancing a Number of Recognizable Kanji
title_sort cross language transfer learning approach via a pretrained preact resnet 18 architecture for improving kanji recognition accuracy and enhancing a number of recognizable kanji
topic deep learning
multiclass classification
language recognition
ResNet
Preact ResNet-18
computer vision
url https://www.mdpi.com/2076-3417/15/9/4894
work_keys_str_mv AT vasylrusyn crosslanguagetransferlearningapproachviaapretrainedpreactresnet18architectureforimprovingkanjirecognitionaccuracyandenhancinganumberofrecognizablekanji
AT andriiboichuk crosslanguagetransferlearningapproachviaapretrainedpreactresnet18architectureforimprovingkanjirecognitionaccuracyandenhancinganumberofrecognizablekanji
AT lesiamochurad crosslanguagetransferlearningapproachviaapretrainedpreactresnet18architectureforimprovingkanjirecognitionaccuracyandenhancinganumberofrecognizablekanji