Attempts to recognize anomalously deformed Kana in Japanese historical documents

Nguyen, Hung Tuan; Ly, Nam Tuan; Nguyen, Kha Cong; Nguyen, Cuong Tuan; Nakagawa, Masaki

doi:10.1145/3151509.3151514

Cited by 29 publications

(11 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This approach is very computationally appealing, but is inappropriate for the general Kuzushiji recognition task due to the contextual nature of many characters. This was explored by [16], where they used a dataset for the paper in which the problem was simplified with input given as a single column of already extracted kuzushiji characters.…”

Section: Work Prior To Kuronetmentioning

confidence: 99%

KuroNet: Regularized Residual U-Nets for End-to-End Kuzushiji Character Recognition

2020

View full text Add to dashboard Cite

Kuzushiji, a cursive writing style, had been used in Japan for over a thousand years starting from the eighth century. Over 3 million books on a diverse array of topics, such as literature, science, mathematics and even cooking are preserved. However, following a change to the Japanese writing system in 1900, Kuzushiji has not been included in regular school curricula. Therefore, most Japanese natives nowadays cannot read books written or printed just 150 years ago. Museums and libraries have invested a great deal of effort into creating digital copies of these historical documents as a safeguard against fires, earthquakes and tsunamis. The result has been datasets with hundreds of millions of photographs of historical documents which can only be read by a small number of specially trained experts. Thus there has been a great deal of interest in using machine learning to automatically recognize these historical texts and transcribe them into modern Japanese characters. Our proposed model KuroNet (which builds on Clanuwat et al. in International conference on document analysis and recognition (ICDAR), 2019) outperforms other model for Kuzushiji recognition. In this paper, KuroNet achieves higher accuracy while still recognizing entire pages of text using the residual U-Net architecture from adding more regularization. We also explore areas where our system is limited and suggests directions for future work.

show abstract

Section: Work Prior To Kuronetmentioning

confidence: 99%

KuroNet: Regularized Residual U-Nets for End-to-End Kuzushiji Character Recognition

2020

View full text Add to dashboard Cite

show abstract

“…This approach is very computationally appealing, but is inappropriate for the general Kuzushiji recognition task due the contextual nature of many characters. This was explored by Nguyen et al (2017). Clanuwat et al (2018a) produced datasets consisting of individually segmented kuzushiji characters, but did not consider the Kuzushiji recognition task in general.…”

Section: Related Workmentioning

confidence: 99%

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

Clanuwat

Lamb

Kitamoto

2019

2019 International Conference on Document Analysis and Recognition (ICDAR)

View full text Add to dashboard Cite

Kuzushiji, a cursive writing style, had been used in Japan for over a thousand years starting from the 8th century. Over 3 millions books on a diverse array of topics, such as literature, science, mathematics and even cooking are preserved. However, following a change to the Japanese writing system in 1900, Kuzushiji has not been included in regular school curricula. Therefore, most Japanese natives nowadays cannot read books written or printed just 150 years ago. Museums and libraries have invested a great deal of effort into creating digital copies of these historical documents as a safeguard against fires, earthquakes and tsunamis. The result has been datasets with hundreds of millions of photographs of historical documents which can only be read by a small number of specially trained experts. Thus there has been a great deal of interest in using Machine Learning to automatically recognize these historical texts and transcribe them into modern Japanese characters. Nevertheless, several challenges in Kuzushiji recognition have made the performance of existing systems extremely poor. To tackle these challenges, we propose KuroNet, a new end-to-end model which jointly recognizes an entire page of text by using a residual U-Net architecture which predicts the location and identity of all characters given a page of text (without any preprocessing). This allows the model to handle long range context, large vocabularies, and non-standardized character layouts. We demonstrate that our system is able to successfully recognize a large fraction of pre-modern Japanese documents, but also explore areas where our system is limited and suggest directions for future work.

show abstract

“…Nguyen et al won the competition. They developed three recognition systems based on convolutional neural network (CNN) and Bidirectional Long Short-Term Memory (BLSTM) for three tasks [5]. For recognizing isolated Kuzushiji characters, they employed CNN and 2DBLSTM based methods.…”

Section: For Recognizing Kuzushiji Characters Horiuchi Andmentioning

confidence: 99%

A Human-Inspired Recognition System for Pre-Modern Japanese Historical Documents

Clanuwat

Kitamoto

2019

IEEE Access

View full text Add to dashboard Cite

Recognition of historical documents is a challenging problem due to the noised, damaged characters, and background. However, in Japanese historical documents, not only contains the mentioned problems, pre-modern Japanese characters were written in cursive and are connected. Therefore, character segmentation-based methods do not work well. This leads to the idea of creating a new recognition system. In this paper, we propose a human-inspired document reading system to recognize multiple lines of premodern Japanese historical documents. During the reading, people employ eyes movement to determine the start of a text line. Then, they move the eyes from the current character/word to the next character/word. They can also determine the end of a line or skip a figure to move to the next line. The eyes movement integrates with visual processing to operate the reading process in the brain. We employ attention-based encoder-decoder to implement this recognition system. First, the recognition system detects were to start a text line. Second, the system scans and recognize character by character until the text line is completed. Then, the system continues to detect the start of the next text line. This process is repeated until reading the whole document. As results, the system is successful to recognize multiple lines, connected and cursive characters without performing character/line segmentation. Besides, we also employ a coverage model which stores the history of eyes movement to predict the next movement more precisely. We tested our human-inspired recognition system on the pre-modern Japanese historical document provided by the PRMU Kuzushiji competition. The results of the experiments demonstrate the superiority and effectiveness of our proposed system by achieving Sequence Error Rate of 9.87% and 53.81% on level 2 and level 3 of the dataset, respectively. These results outperform to any other systems participated in the PRMU Kuzushiji competition. INDEX TERMS A human reading-inspired recognition system, recognition of pre-modern Japanese historical document, attention-based encoder-decoder, Kuzushiji.

show abstract

Attempts to recognize anomalously deformed Kana in Japanese historical documents

Cited by 29 publications

References 11 publications

KuroNet: Regularized Residual U-Nets for End-to-End Kuzushiji Character Recognition

KuroNet: Regularized Residual U-Nets for End-to-End Kuzushiji Character Recognition

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

A Human-Inspired Recognition System for Pre-Modern Japanese Historical Documents

Contact Info

Product

Resources

About